UMass Lowell/Boston Globe Massachusetts Statewide Poll (October 2018) · Oct 2018 – Oct 7, 2018 · US: Massachusetts
Methodology
- Pollster
- SSRS academic
- Sponsor
- UMass Lowell Center for Public Opinion / The Boston Globe
- Commissioned by
- The Boston Globe
- Fielded
- Oct 2018 – Oct 7, 2018
- Population
- registered voters
- Sample size
- 791
- Sample frame
- probability RDD
- Mode
- telephone - dual-frame RDD, live interviewer; 433 landline and 519 cell phone interviews (275 cell-only), up to five contact attempts per number
- Margin of error
- +/-4.4 percentage points for registered voters (+/-3.9 for all adults), design-effect adjusted
- Weighting
- base weight for dual-frame overlap, then raking (iterative proportional fitting) to 2018 CPS state parameters on sex by age, sex by education, age by education, race and state region, plus telephone-use status from the 2012-2016 National Health Interview Survey
- Geography
- US: Massachusetts
- Base notes
- 791 registered voters drawn from 952 Massachusetts adults interviewed; a 485-case likely-voter subsample is reported alongside the RV figures and the crosstabs are of likely voters
Sources
Verification note. Topline and SSRS methods report fetched directly from uml.edu; wording, percentages, party breaks, n, field dates, design effect and MOE all read off the PDFs. The design effect is 1.55 overall and 1.58 for registered voters. BALLOT BENCHMARK: Massachusetts Question 3, voted 2018-11-06, certified Yes (keep the law) 1,806,742 = 67.82%, No (repeal) 857,401 = 32.18%. The ballot test here (RV 73/21, LV 74/22) overstated the Yes share by 5-6 points. The most valuable feature of this poll is internal: the SAME respondents who split 74/22 for Yes on the ballot wording split only 56/33 for allowing gender-identity restroom use when the underlying policy was described directly (Q10). That roughly 18-point gap between the ballot frame and the policy frame, inside one instrument, is the cleanest available measurement of how much of Question 3's margin came from the anti-discrimination framing rather than from support for the underlying access rule - and it points at the same mechanism that produced the late break toward repeal.
Figures below are transcribed from these documents. Where they disagree with the source, the source governs.
Massachusetts Question 3 ballot test - gender-identity public accommodations veto referendum
Question 3 asks if voters want to keep or repeal a recently passed Massachusetts law that adds gender identity to the list of prohibited grounds for discrimination in places of public accommodation, resort, or amusement. If the election was being held today, how would you vote, YES to keep in place the current law, which prohibits discrimination on the basis of gender identity in places of public accommodation, or NO to repeal this provision of the public accommodation law?
Yes (keep the law)
73%
No (repeal)
21%
Don't Know/Unsure/None
5%
Refused
1%
Wording notes: tracks the official ballot summary language closely, including 'resort, or amusement' | describes the measure purely as a discrimination question, never mentioning restrooms or locker rooms | undecideds pushed with a follow-up leaner (Q9a)
neutral — closely mirrors the measure's official summary and states both vote consequences; it is not slanted relative to the ballot, but because the ballot language itself omits facility access it measures something narrower than the policy at issue - which is precisely what the paired Q10 exposes
Subgroup figures (16)
restroom policy for transgender people, asked of the same sample immediately after the Question 3 ballot test
Question 3 affects transgender individuals' use of public bathrooms, fitting rooms and locker rooms, as well as other public spaces. In terms of policies governing public restrooms, do you think these policies should - (require transgender individuals to use the restroom that corresponds with their birth gender) or should these policies (allow transgender individuals to use the restroom that corresponds with their gender identity)?
Allow to use gender identity restroom
54%
Require transgender individuals to use birth gender restroom
34%
Don't Know (VOL)
9%
Refused
2%
Wording notes: uses 'birth gender' | the two alternatives are stated in parallel and were bracketed for rotation | explicitly ties the item back to Question 3, so it is a within-instrument framing comparison rather than an independent measurement
neutral — a symmetric forced choice with both policies spelled out; 'birth gender' is contested but is matched against 'gender identity' rather than standing alone
Subgroup figures (16)