Results of SurveyUSA Election Poll #15972
 
Washington State Initiative 1033: Upside Down I'm Turning You: Support for Initiative 1033 has fallen, 7 days until ballots are counted, according to a SurveyUSA poll conducted for KING-TV Seattle. 3 weeks ago, SurveyUSA showed 1033 backed 4:3. Today, SurveyUSA shows 1033 opposed 5:4. Here are important methodological considerations:

In SurveyUSA's 10/06/09 poll, respondents were read a SurveyUSA-written summary of the 1033 language that appears on the ballot. SurveyUSA reported the measure ahead 45% to 32%. Some who did not agree with those poll results argued that if SurveyUSA had read actual ballot language to respondents, instead of a summary, the poll results would have been different.

For today's 10/27/09 SurveyUSA release, a "split ballot" design was used, where half of respondents chosen at random were read the exact ballot language, and the other half were read the same summary that SurveyUSA used in its 10/06/09 poll. There are two separate pieces of learning that come out of the split-ballot:

1 : Respondents who hear the exact ballot language vote the same way as respondents who hear a summary.
2: There is materially less support for 1033 than there was on 10/06/09.

Of those who heard the summary, 39% are certain to vote 'Yes,' 49% are certain to vote 'No.' Of those today who heard the exact ballot language, 38% are certain to vote 'Yes,' 51% are certain to vote 'No.' When these populations are combined (as they are in Question #3 below), the outcome is: 38% certain 'Yes,' 50% certain 'No,' 12% not yet certain. 40% of voters interviewed by SurveyUSA have already voted. 1033 trails among those who have already voted and among those who have not yet voted but who promise to do so before the deadline. 100% of voting is by US mail.

Caveat: There are limitations to polling in off-year elections, limitations to polling on ballot measures and limitations to polling in a world in which many respondents decline to be interviewed and other potential respondents are not reachable. Within those understood limitations, these results permit an observer to see movement from one survey to the next. These results should not be marshaled as "proof" that 1033 will pass or fail when ballots are counted. Those with a vested interest in the outcome may argue that the measure is trailing by more than SurveyUSA shows, or by less than SurveyUSA shows. But what is not possible to argue, is that there is less support for the measure today than there was 3 weeks ago.

 
Referendum 71 today is favored 7:6, which is slightly more support than Referendum 71 had on 10/06/09. Women increasingly support 71, and the more women who vote, the more likely 71 is to pass. Still, the poll results are too even to permit anything definitive to be said about the outcome when 71 votes are counted next week. Younger voters support, seniors oppose. Greater Seattle and Western WA support. Eastern WA opposes. Support is up among the more affluent. Opposition is up among the less affluent.
 
Filtering: 900 state of Washington adults were interviewed by SurveyUSA 10/24/09 through 10/26/09. Of them, 764 were registered to vote. Of them, 252 told SurveyUSA they had already voted and another 336 were determined by SurveyUSA to be almost certain to return a ballot before the deadline on 11/03/09. Registered voters were randomly assigned by SurveyUSA software, in real-time, to hear one of the two versions of Initiative 1033.
 
Summary of ballot language read to 1/2 of respondents: "Initiative 1033 would limit spending for state, county and local governments. On Initiative 1033, are you... Certain to vote yes? Certain to vote no? Or not certain?" {Voters who were not certain were asked: At this hour, on Initiative 1033, do you ... lean toward yes? lean toward no, or do you not lean?}
274 Split Sample Of Likely & Actual VotersAllGenderAge<50 / 50+RaceParty AffiliationIdeologyCollege GradIncomeRegionAlready Voted?
Margin of Sampling Error: ± 6%MaleFemale18-3435-4950-6465+18-4950+WhiteBlackHispanicAsian/OtRepublicDemocratIndependConservaModerateLiberalYesNo< $50K> $50KWestern Eastern Metro SeYesNo
Certain Yes39%47%32%41%40%43%32%40%38%38%****47%57%27%43%56%37%18%42%36%37%40%46%47%32%44%36%
Certain No49%43%54%41%48%50%54%46%52%50%****42%33%63%41%34%53%68%49%50%47%51%45%42%54%51%47%
Not Certain12%10%14%18%13%7%14%14%10%12%****11%10%10%15%10%10%14%9%14%16%10%9%11%14%5%16%
Total100%100%100%100%100%100%100%100%100%100%100%100%100%100%100%100%100%100%100%100%100%100%100%100%100%100%100%100%
Composition of Split Sample Of Likely & Actual Voters100%49%51%15%32%30%23%47%53%86%2%3%8%23%44%32%30%44%21%51%49%30%70%28%22%50%38%62%
 
 
2Exact ballot language read to 1/2 of respondents: "Initiative Measure No. 1033 concerns state, county and city revenue. This measure would limit growth of certain state, county and city revenue to annual inflation and population growth, not including voter-approved revenue increases. Revenue collected above the limit would reduce property tax levies. On Initiative 1033, are you... Certain to vote yes? Certain to vote no? Or not certain?" {Voters who were not certain were asked: At this hour, on Initiative 1033, do you ... lean toward yes? lean toward no, or do you not lean?}
287 Split Sample Of Likely & Actual VotersAllGenderAge<50 / 50+RaceParty AffiliationIdeologyCollege GradIncomeRegionAlready Voted?
Margin of Sampling Error: ± 5.9%MaleFemale18-3435-4950-6465+18-4950+WhiteBlackHispanicAsian/OtRepublicDemocratIndependConservaModerateLiberalYesNo< $50K> $50KWestern Eastern Metro SeYesNo
Certain Yes38%44%31%45%34%36%38%38%37%36%****44%41%29%45%53%33%24%35%40%37%37%42%47%32%40%36%
Certain No51%49%53%39%55%52%53%49%53%52%****54%43%61%45%37%54%67%56%47%52%51%48%43%55%54%48%
Not Certain12%7%16%17%11%12%8%13%10%13%****3%16%11%10%10%13%8%9%14%11%11%10%10%13%6%16%
Total100%100%100%100%100%100%100%100%100%100%100%100%100%100%100%100%100%100%100%100%100%100%100%100%100%100%100%100%
Composition of Split Sample Of Likely & Actual Voters100%51%49%20%32%28%21%51%49%86%2%4%8%27%40%31%33%43%20%49%51%33%67%25%23%52%42%58%
 
 
3Question 3 shows the poll results when the split-samples from Question 1 above and Question 2 above are combined. Question 3 contains the poll results that should be used when media report on this survey.
561 Likely & Actual VotersAllGenderAge<50 / 50+RaceParty AffiliationIdeologyCollege GradIncomeRegionAlready Voted?
Margin of Sampling Error: ± 4.2%MaleFemale18-3435-4950-6465+18-4950+WhiteBlackHispanicAsian/OtRepublicDemocratIndependConservaModerateLiberalYesNo< $50K> $50KWestern Eastern Metro SeYesNo
Certain Yes38%45%31%43%37%39%35%39%38%37%****45%48%28%44%55%35%21%38%38%37%38%44%47%32%42%36%
Certain No50%46%53%40%51%51%54%47%52%51%****48%39%62%43%36%53%68%52%48%50%51%46%43%55%53%48%
Not Certain12%9%15%17%12%10%11%14%10%12%****7%13%10%13%10%12%11%9%14%13%11%10%10%14%5%16%
Total100%100%100%100%100%100%100%100%100%100%100%100%100%100%100%100%100%100%100%100%100%100%100%100%100%100%100%100%
Composition of Likely & Actual Voters100%50%50%17%32%29%22%49%51%86%2%4%8%25%42%32%32%43%20%50%50%32%68%26%23%51%40%60%
 
 
Referendum 71 would expand the rights, responsibilities, and obligations of registered domestic partners and their families. On Referendum 71, are you... Certain to vote to approve? Certain to vote to reject? Or not certain? {Voters who were not certain were asked: At this hour, on Referendum 71, do you ... lean toward voting to approve? lean toward voting to reject, or do you not lean?}
561 Likely & Actual VotersAllGenderAge<50 / 50+RaceParty AffiliationIdeologyCollege GradIncomeRegionAlready Voted?
Margin of Sampling Error: ± 4.2%MaleFemale18-3435-4950-6465+18-4950+WhiteBlackHispanicAsian/OtRepublicDemocratIndependConservaModerateLiberalYesNo< $50K> $50KWestern Eastern Metro SeYesNo
Certain Approve50%44%55%56%49%52%43%51%48%50%****62%21%73%43%21%59%77%53%48%44%54%52%35%55%53%47%
Certain Reject43%48%37%37%43%43%48%41%45%43%****32%73%20%48%74%33%14%42%43%46%41%44%54%37%42%43%
Not Certain7%7%7%8%8%5%9%8%7%8%****6%6%7%9%5%8%9%5%9%10%6%4%11%7%4%9%
Total100%100%100%100%100%100%100%100%100%100%100%100%100%100%100%100%100%100%100%100%100%100%100%100%100%100%100%100%
Composition of Likely & Actual Voters100%50%50%17%32%29%22%49%51%86%2%4%8%25%42%32%32%43%20%50%50%32%68%26%23%51%40%60%
 
 
** Too few respondents of this type were interviewed for this data to be meaningful.