September 18, 2018
Trump at 42 percent approval, voters divided on how he has changed GOP
Please note: Complete poll results and methodology information can be found online at
MILWAUKEE — A new Marquette Law School Poll of Wisconsin voters finds Democrat Tony Evers with 49 percent support and incumbent Republican Scott Walker with 44 percent support among likely voters in the state's race for governor. Libertarian candidate Phil Anderson receives 6 percent. Among likely voters, those who say they are certain to vote in the November election, only 1 percent say they lack a preference or do not lean to a candidate. In the previous Marquette Law School Poll conducted in August, the race for governor was tied among likely voters, with Evers at 46 percent, Walker at 46 percent and Anderson with 6 percent.
In the race for U.S. Senate, incumbent Democrat Tammy Baldwin leads among likely voters with 53 percent, while 42 percent support Republican Leah Vukmir and 4 percent say they lack a preference or do not lean toward a candidate. In August, the race was closer, with Baldwin receiving 49 percent and Vukmir 47 percent.
In the race for attorney general, incumbent Republican Brad Schimel is the choice of 48 percent and Democrat Josh Kaul is the choice of 41 percent of likely voters. Ten percent lack a preference in this race. This is the first time this year's attorney general race has been measured by the Marquette poll.
Among all registered voters surveyed in the poll, Evers receives 47 percent support, Walker receives 43 percent and Anderson 7 percent in the race for governor.
Among all registered voters in the Senate race, Baldwin receives 52 percent support and Vukmir 40 percent.
For attorney general, registered voters give Schimel 47 percent support and Kaul 40 percent.
The poll was conducted Sept. 12-16, 2018. The sample includes 800 registered voters in Wisconsin interviewed by cell phone or landline, with a margin of error of +/- 4 percentage points. For likely voters, the sample size is 614 and the margin of error is +/- 4.4 percentage points.
Eight issue questions were asked of half the sample. The state issues have a sample size of 406 and a margin of error of +/- 5.7 percentage points. The national issues have a sample size of 394 and a margin of error of +/- 5.7 percentage points. The half-sample items are listed at the end of this release.
The non-incumbent candidates are substantially less well-known than the incumbents, though voters are becoming more familiar with the candidates over time.
Awareness of Evers and Vukmir has increased among likely voters since the Marquette Law School Poll in August. Kaul and Anderson are far behind in name recognition. Results for September and August are shown in Table 1.
Table 1: Favorability and awareness of candidates, Sept. and Aug., among likely voters
| Sept:Fav | Sept:Unfav | Sept:Not Heard/DK | Aug:Fav | Aug:Unfav | Aug:Not Heard/DK | |
| Evers | 40 | 29 | 31 | 38 | 27 | 35 |
| Vukmir | 26 | 38 | 36 | 30 | 29 | 41 |
| Kaul | 7 | 5 | 87 | 4 | 3 | 92 |
| Anderson | 4 | 4 | 92 | 4 | 4 | 91 |
| Walker | 45 | 52 | 2 | 49 | 47 | 4 |
| Baldwin | 48 | 40 | 12 | 46 | 42 | 11 |
| Schimel | 24 | 20 | 56 | 26 | 16 | 57 |
Thirty-one percent of likely voters lack an opinion of Evers, down from 35 percent in August. For Vukmir, 36 percent lack an opinion now, compared to 41 percent in August.
Eighty-seven percent are unable to give an opinion of Kaul in September, compared to 92 percent who could not do so in August.
For Anderson, 92 percent could not give an opinion in September, while 91 percent could not do so in August.
Walker is the best known among all the candidates, with 2 percent lacking an opinion in September and 4 percent without an opinion in August.
Baldwin is not rated by 12 percent in September and by 11 percent in August.
Schimel is much less well-known than the other incumbents, with 56 percent in September unable to rate him and 57 percent in August unable to do so.
Among likely voters, Evers is viewed positively by 40 percent and unfavorably by 29 percent in September. Vukmir is seen favorably by 26 percent and unfavorably by 38 percent. Kaul is seen favorably by 7 percent and unfavorably by 5 percent.
Anderson is seen favorably by 4 percent and unfavorably by 4 percent.
Walker's favorable rating is 45 percent and his unfavorable is 52 percent. For Baldwin, 48 percent have a favorable opinion and 40 percent unfavorable. Schimel is seen favorably by 24 percent with 20 percent unfavorable.
The electorate has become increasingly segmented by gender and education level among white voters. Differences by race have been longer-standing. Table 2 shows preference for governor among white males and females by education, and for non-whites or Hispanic likely voters.
Table 2: Vote for governor by race, gender, and education among likely voters
| White, Male, Noncollege | White, Female, Noncollege | White, Male, College | White, Female, College | Nonwhite or Hispanic | |
| Evers | 40 | 48 | 49 | 62 | 49 |
| Walker | 52 | 46 | 45 | 36 | 37 |
| Anderson | 6 | 4 | 5 | 1 | 12 |
Evers receives his strongest support from white, female college graduates, while Walker does best with white, male non-college graduates.
Partisans are strongly supporting their party's candidate, but independents are currently favoring Evers, as shown in Table 3.
Table 3: Vote for governor by Party ID among likely voters
| Rep | Dem | Ind | |
| Evers | 4 | 93 | 52 |
| Walker | 94 | 3 | 32 |
| Anderson | 2 | 3 | 13 |
Preferences in the Senate race by race, gender, and education are shown in Table 4. Baldwin does best with white, female college graduates but also holds a substantial advantage among non-college white females. White college males are evenly divided, and Vukmir has a substantial advantage with non-college white males.
Table 4: Vote for Senate by race, gender, and education among likely voters
| White, Male, Noncollege | White, Female, Noncollege | White, Male, College | White, Female, College | Nonwhite or Hispanic | |
| Baldwin | 41 | 57 | 48 | 64 | 59 |
| Vukmir | 54 | 40 | 48 | 30 | 34 |
The senate vote by party is shown in Table 5. Partisan alignments are strong, though Vukmir's support among Republicans is not as strong as Baldwin's is among Democrats. Independents favor Baldwin.
Table 5: Vote for senate by party ID among likely voters
| Rep | Dem | Ind | |
| Baldwin | 9 | 97 | 54 |
| Vukmir | 86 | 2 | 38 |
In the attorney general's race, the results are shown in Table 6 and Table 7. Schimel, holds a lead among non-college white males and females but also among white college males and among non-whites. Kaul leads among white college females.
Table 6: Vote for attorney general by race, gender, and education among likely voters
| White, Male, Noncollege | White, Female, Noncollege | White, Male, College | White, Female, College | Nonwhite or Hispanic | |
| Kaul | 35 | 39 | 40 | 50 | 42 |
| Schimel | 51 | 50 | 52 | 40 | 46 |
Table 7: Vote for attorney general by party ID among likely voters
| Rep | Dem | Ind | |
| Kaul | 6 | 80 | 36 |
| Schimel | 90 | 11 | 44 |
Schimel holds 90 percent of Republican support while Kaul receives support from 80 percent of Democrats. Independents lean to Schimel. Both candidates for attorney general are less well-known than the gubernatorial and senate candidates, which may change as the election nears.
Fifty percent of Wisconsin registered voters see the state as headed in the right direction while 47 percent think the state is off on the wrong track. In August, 53 percent said right direction and 41 percent said wrong track. In September 2014, 54 percent said the state was headed in the right direction and 42 percent said it was on the wrong track.
Thirty-one percent think the state budget is in better shape than a few years ago, while 28 percent say it is about the same and 29 percent say the budget is in worse shape. In September 2014, 41 percent said the budget was in better shape, 25 percent said it was about the same and 27 percent said the budget was in worse shape. About twice as many people now, 12 percent, say they don't know how the budget is doing compared to 6 percent in September 2014.
Walker's job approval among registered voters stands at 44 percent, with 50 disapproving. The trend in approval in 2018 is shown in Table 8. Among likely voters, 46 percent approve and 51 percent disapprove. In September 2014, 49 percent approved and 46 percent disapproved among registered voters.
Table 8: Scott Walker job approval trend among registered voters
| Approve | Disapprove | Don't know | |
| September 2018 | 44 | 50 | 5 |
| August 2018 | 48 | 45 | 6 |
| July 2018 | 47 | 45 | 7 |
| June 2018 | 49 | 47 | 3 |
A majority of registered voters are satisfied with the job public schools are doing in their community, with 18 percent saying they are very satisfied and 46 percent saying satisfied. Seventeen percent say they are dissatisfied and 8 percent are very dissatisfied.
Satisfaction with public schools is the lowest in the city of Milwaukee, where 46 percent say they are either very satisfied or satisfied while 48 percent are dissatisfied or very dissatisfied. In each of the other four regions of the state, more than 60 percent say they are very satisfied or satisfied while less than 30 percent in each region are dissatisfied or very dissatisfied. Results by region are shown in Table 9.
Table 9: School satisfaction by region among registered voters
| MKE City | Rest of MKE area | Madison | Green Bay | North & West | |
| Very satisfied | 8 | 18 | 17 | 14 | 23 |
| Satisfied | 38 | 43 | 47 | 50 | 49 |
| Dissatisfied | 29 | 19 | 14 | 18 | 13 |
| Very dissatisfied | 19 | 8 | 11 | 4 | 5 |
Voters say they would rather increase spending on public schools than reduce property taxes by a 57 percent to 38 percent margin. Support for additional spending on public schools has increased since the question was first asked in 2013. The full trend on this issue is shown in Table 10.
Table 10: Trend in property tax vs school spending opinion, 2013-2018 among registered voters
| Cut property taxes | Increase school spending | |
| 3/11-13/13 | 49 | 46 |
| 5/6-9/13 | 49 | 46 |
| 4/7-10/15 | 40 | 54 |
| 2/25-3/1/18 | 33 | 63 |
| 6/13-17/18 | 35 | 59 |
| 8/15-19/18 | 32 | 61 |
| 9/12-16/18 | 38 | 57 |
Five percent of registered voters say roads where they live are in excellent condition, 30 percent say good condition, 32 percent say fair condition, and 32 percent say poor condition.
There are differences between the southern and northern parts of the state in opinion of roads, with the more negative views in the southern part of the state. Note: A data coding error in the July poll reversed the results by region. The corrected results for July are shown in Table 11. Table 12 gives the results for the September poll.
Table 11: Condition of roads by region in July poll (corrected) among registered voters
| South | North | |
| Excellent | 6 | 3 |
| Good | 28 | 45 |
| Fair | 41 | 30 |
| Poor | 25 | 22 |
Table 12: Condition of roads by region in September poll among registered voters
| South | North | |
| Excellent | 4 | 7 |
| Good | 26 | 35 |
| Fair | 34 | 29 |
| Poor | 36 | 28 |
Forty-eight percent of registered voters think the state is paying more than the Foxconn plant is worth, while 39 percent think the plant will provide at least as much value as the state is investing in the plant. Thirteen percent say they don't know if the plant will be worth it or not. In the August poll, 44 percent said the state was paying too much and 41 percent said it was worth it.
A 58 percent majority of registered voters statewide believe the Foxconn plant will substantially improve the economy of the greater Milwaukee area, while 31 percent do not think it will and 11 percent say they don't know. In the August poll, 61 percent said the Milwaukee area would benefit while 27 percent did not think so.
When asked if businesses where the respondent lives will benefit from Foxconn, 33 percent say businesses will benefit directly from the Foxconn plant, while 54 percent say their local businesses will not benefit and 14 percent don't know. In the August poll 32 percent said their local businesses would benefit, while 61 percent did not think so.
Fifty-nine percent say that protection of the environment should be given priority even at the risk of curbing economic growth. Twenty-seven percent would give economic growth priority even if the environment suffers to some extent.
Forty-seven percent of registered voters say they have a favorable view of the Affordable Care Act, also called Obamacare, while 45 percent have an unfavorable view.
Asked what should be done about the ACA/Obamacare health reform, 4 percent say keep the law as it is, 55 percent say keep and improve the law, 25 percent say repeal and replace the health reform law and 10 percent say the law should be repealed and not replaced. Opinion on the health reform law has changed little since June 2017, when 6 percent said keep as the law, 54 percent said improve it, 27 percent said repeal and replace and 7 percent said repeal and not replace the law.
U.S. Supreme Court nominee Brett Kavanaugh has a 29 percent favorable rating while 29 percent have an unfavorable view and 41 percent are not able to give a rating. In July, when last asked, 27 percent gave a favorable rating while 22 percent had an unfavorable view and 50 percent were unable to rate him. All but 152 voter interviews for the poll were completed before Sunday, when details emerged of an alleged sexual assault when Kavanaugh was in high school.
Thirty-one percent think increased tariffs on steel and aluminum imports will improve the U.S. economy while 52 percent think tariffs will hurt the economy. Seventeen percent say they don't know. In the August poll, 34 percent said tariffs would improve the economy while 48 percent said tariffs would hurt the economy.
Partisan differences in views of tariffs are sharp, with 59 percent of Republicans saying tariffs are good for the U.S. economy, 20 percent saying they are bad for the economy and 20 saying they don't know. A large majority of Democrats say tariffs are bad for the economy and a majority of independents see tariffs as a bad thing for the economy, as shown in Table 13.
Table 13: View of tariffs by Party ID among registered voters
| Rep | Dem | Ind | |
| Good for Econ | 59 | 8 | 28 |
| Bad for Econ | 20 | 76 | 56 |
| DK | 20 | 16 | 16 |
On free trade agreements in general, 58 percent think these agreements have been a good thing for the U.S. economy, while 25 percent think they have been bad for the economy. Sixteen percent say they don't know. In August, 55 percent said free trade has been good for the economy while 29 percent said free trade has been bad for the economy, while fourteen percent said they didn't know.
Views of abortion remain close to their long-term averages in polling since 2012, with 26 percent saying abortion should be legal in all cases, 36 percent saying legal in most cases, 21 percent illegal in most cases and 9 percent saying it should be illegal in all cases. The long-term average over eight statewide polls is 26 percent legal in all cases, 34 percent legal in most, 23 percent illegal in most and 12 percent saying illegal in all cases.
There has been an increase in support for a path to citizenship for undocumented immigrants in the years since 2012. Sixty-eight percent say undocumented immigrants should be allowed to stay in the U.S. and eventually apply for citizenship, 15 percent say undocumented immigrants should be allowed to stay as temporary guest workers but with no path to citizenship and 14 percent say such immigrants should be required to leave the country. When first asked in September 2012, 51 percent supported a path to citizenship, 23 percent favored a guest worker status and 20 percent preferred requiring undocumented immigrants to leave.
Support for a path to citizenship has increased across each party identification group. Table 14 shows preferences on this issue by party among registered voters in the current poll, and Table 15 shows how partisans divided in September 2012.
Table 14: Policy for undocumented immigrants by party ID September 2018 among registered voters
| Rep | Dem | Ind | |
| Path to citizenship | 43 | 90 | 69 |
| Guest worker | 26 | 7 | 14 |
| Leave US | 27 | 0 | 16 |
| DK | 4 | 3 | 1 |
Table 15: Policy for undocumented immigrants by party ID September 2012 among registered voters
| Rep | Dem | Ind | |
| Path to citizenship | 36 | 64 | 51 |
| Guest worker | 29 | 19 | 24 |
| Leave US | 31 | 14 | 19 |
| DK | 3 | 4 | 4 |
In the new poll results, President Donald Trump has a 42 percent approval rating, with 54 percent disapproving. In the previous Marquette Law School poll in August, his approval was 45 percent with 51 percent disapproving. Partisans are deeply divided on Trump's job performance, as shown in Table 16.
Table 16: Trump job approval by party ID among registered voters
| Rep | Dem | Ind | |
| Approve | 81 | 4 | 41 |
| Disapprove | 14 | 94 | 54 |
| DK | 4 | 1 | 4 |
Asked if Trump has changed the Republican party — and if so, how — 23 percent say he has changed it for the better, 49 percent say he has changed it for the worse and 23 percent say he hasn't changed it much either way.
Partisans have differing views of Trump's effect on his party, as seen in Table 17.
Table 17: How Trump has changed GOP by party ID among registered voters
| Rep | Dem | Ind | |
| For better | 51 | 3 | 17 |
| For worse | 13 | 85 | 49 |
| Not changed | 29 | 11 | 31 |
| DK | 6 | 2 | 3 |
There are some divisions among Republican voters. Asked if they had a favorable or unfavorable view of the late Sen. John McCain, among all registered voters, 70 percent said favorable while 19 percent had an unfavorable view. When broken down by party identification in Table 18, more Democrats and independents had a favorable view of the 2008 Republican presidential nominee than did Republicans.
Table 18: Opinion of John McCain by party ID among registered voters
| Rep | Dem | Ind | |
| Favorable | 59 | 85 | 65 |
| Unfavorable | 28 | 10 | 20 |
| Not heard/DK | 11 | 5 | 15 |
Among Republicans with a favorable view of McCain, Trump's approval rating is 72 percent while among those with an unfavorable view of McCain it is 93 percent. A smaller difference also appears for approval of Walker. Among those Republicans with a favorable view of McCain, Walker's approval rating is 85 percent, while among those with an unfavorable view of McCain it is 94 percent.
Overall, 62 percent of registered voters say they are very enthusiastic about voting in this year's elections, with 24 percent somewhat enthusiastic and 13 percent either not very or not at all enthusiastic.
Among Republicans, 64 percent are very enthusiastic, while among Democrats 75 percent are. Among independents, 49 percent say they are very enthusiastic about voting this year. In August, 69 percent of Republicans and 67 percent of Democrats were very enthusiastic, along with 56 percent of independents.
The Marquette Law School Poll is the most extensive statewide polling project in Wisconsin history. This poll interviewed 800 registered Wisconsin voters by landline or cell phone Sept. 12-16, 2018. The margin of error is +/-4 percentage points for the full sample.
For likely voters, the sample size is 614 and the margin of error is +/- 4.4 percentage points.
Eight issue questions were asked of half the sample. The state issues have a sample size of 406 and a margin of error of +/- 5.7 percentage points. The national issues have a sample size of 394 and a margin of error of +/- 5.7 percentage points. State half-sample issues include Foxconn (3 items) and priority of the environment or economic growth. Half-sample national issues involve tariffs and free trade, how to deal with undocumented immigrants and preference for abortion policy.
The partisan makeup of the sample, including those who lean to a party, is 45 percent Republican, 46 percent Democratic and 8 percent independent. The long-term total for the previous 47 statewide Marquette polls, with 41,752 respondents, is 43 percent Republican and 47 percent Democratic, with 9 percent independent. The partisan makeup of the current sample, excluding those who lean to a party, is 32 percent Republican, 33 percent Democratic and 34 percent independent, compared to the long-term totals of 27 percent Republican, 31 percent Democratic and 41 percent independent.
The entire questionnaire, methodology statement, full results and breakdowns by demographic groups are available at
