Should we trust the Gavin Newsom recall polls showing an easy win? - SFGate

Two separate polls of the Gavin Newsom recall election show the governor opening up large leads in the final days of the campaign.

Of course, polls — especially ones from 2016 and beyond — that show Democrats with comfortable leads have not always been accurate. The 2020 presidential election featured a polling miss as large as the 2016 miss, and a great deal has been written on why polls seem to consistently undersample conservative voters.

Rather than enter a discussion around "shy" Trump/recall voters, response rates in the work-from-home era or weighting by education, perhaps the best way to determine the reliability of the Newsom recall polls is by seeing how well Survey USA and PPIC performed in California's last statewide election — which was less than a year ago with pollsters working under the same COVID-19 related constraints.

Below is a table showing numbers from Survey USA's final poll for the 2020 general election compared with actual results. Included are the results of four races: 1. The presidential race between Joe Biden and Donald Trump, 2. California's Proposition 15 (raising property taxes on commercial and industrial properties), 3. California's Proposition 16 (lifting a ban on affirmative action) and 4. California's Proposition 22 (granting app-based transportation and delivery companies an exception to a state labor law by classifying drivers as independent contractors rather than employees).

The "conservative" side for the three ballot measures are "no" on Prop. 15, "no" on Prop. 16 and "yes" on Prop. 22. Because a recall election is a special election with lower-than-expected turnout, looking at past polling on ballot measures may be more useful than polling on big-ticket, top-of-ballot races like the presidential election.

Survey USA logged some massive misses on Prop. 15 and Prop. 16, largely explained by the large portion of respondents who said they were undecided. On Prop. 15, Survey USA found 49% of respondents voting "yes," 21% voting "no" and a whopping 30% saying they were undecided. In the end, 52% of Californians voted "no" and 48% voted "yes," so unless every single undecided voter decided to vote "no" at the last second, this constitutes a major miss.

For Prop. 16, Survey USA's polling found 40% of respondents voting "yes," 26% voting "no" and 34% undecided. The final result was 57% voting "no" and 43% voting "yes." Survey USA's polling for Prop. 22, on the other hand, was fairly accurate, predicting a win of 14 percentage points (Prop. 22 passed by 17 percentage points). If we work under the assumption that Survey USA's undecided voters consistently seem to end up taking the "conservative" side, then the recall could be a nail-biter, as Survey USA's final recall survey shows 51% opposing the recall, 43% in support and 6% undecided.

PPIC fared better than Survey USA in 2020 — but still missed on Prop. 15 and did not ask respondents about Prop. 22.

If we take PPIC's worst miss from 2020 (Prop. 15) and assume a similar eight-percentage-point miss in the recall, Newsom would still survive by 11 percentage points, as PPIC's final recall poll shows Newsom leading by 19.

Polls from earlier in the summer attempted to weight by how likely a voter is to actually cast a ballot, as Republicans have been more enthusiastic about the recall than Democrats. More recent polls show the enthusiasm gap shrinking, and given the makeup of California's electorate (47% Democratic, 24% Republican, 23% nonaffiliated, 6% other), that's very good news for Newsom.

Additional polling, including from the Berkeley Institute of Government Studies, is expected ahead of the Sept. 14 recall election.

Adblock test (Why?)

"easy" - Google News
September 08, 2021 at 06:04PM
https://ift.tt/38Q3u1R

Should we trust the Gavin Newsom recall polls showing an easy win? - SFGate
"easy" - Google News
https://ift.tt/38z63U6
Shoes Man Tutorial
Pos News Update
Meme Update
Korean Entertainment News
Japan News Update

Bagikan Berita Ini