09/03/2016 12:40 AM AEDT | Updated 21/12/2016 4:39 AM AEDT

HUFFPOLLSTER: Ted Cruz And John Kasich Battle For Second Place In Michigan

Donald Trump leads in the Great Lakes State,while Marco Rubio has fallen to the bottom of the pack.

Jim Young / Reuters

Hillary Clinton and Donald Trump lead in Michigan, but John Kasich could provide a challenge for Trump. If Trump doesn’t win in Ohio and Florida next week, the GOP could be headed for a contested convention. And there's reason to be skeptical of forecasts that show Republicans strong in the general election. This is HuffPollster for Tuesday, March 8, 2016.

DONALD TRUMP LEADS IN MICHIGAN, BUT WATCH OUT FOR KASICH - Voters in Michigan and Mississippi head to the polls on Tuesday, as do Republicans in Hawaii in Idaho. Michigan is by far the largest prize, with 147 delegates at stake for Democrats and 59 delegates to be awarded to Republicans. Polls of Michigan Democrats show Hillary Clinton leading Sen. Bernie Sanders (I-Vt.) by margins ranging from 11 to 37 points. The Republican polls show Donald Trump in the lead, but with a race for second place between Sen. Ted Cruz (R-Texas) and Ohio Gov. John Kasich. Sen. Marco Rubio (R-Fla.) has fallen into fourth place in many  recent surveys.

Pollsters find evidence of Kasich surging - From a Monmouth University Poll press release: “‘After this past weekend’s mixed bag of results, Trump appears positioned for a win in Michigan, but the race may be tightening in the final hours,’ said Patrick Murray, director of the independent Monmouth University Polling Institute. ‘Trump’s support may be dropping, while Kasich’s star could be rising.’ In interviews conducted Thursday and Friday, Trump held a solid lead with 39% support compared to 22% for Cruz, 17% for Kasich, and 14% for Rubio. The race was much tighter in Saturday and Sunday interviews at 32% for Trump, 26% for Kasich, 25% for Cruz, and 12% for Rubio.” [Monmouth]

Kasich does especially well in some surveys - Across primaries so far, John Kasich has consistently done better in polls conducted by the American Research Group than in most other polls, including a Michigan poll released this week that showed Kasich leading Trump by two points. ARG’s pollster Dick Bennett told HuffPollster that Kasich’s strength seemed to be largely among respondents reached who had to be contacted more than once before they answered the survey: “We looked in New Hampshire and South Carolina and found that Kasich was stronger in our callbacks….In Michigan, Kasich was also stronger in the callbacks in the first survey, but, for the first time, he was stronger in initial contacts in the latest survey. We shall see what happens to that.”

HUFFPOLLSTER VIA EMAIL! - You can receive this daily update every weekday morning via email! Justclick here, enter your email address, and click "sign up." That's all there is to it (and you can unsubscribe anytime).

FLORIDA AND OHIO ARE KEY TO TRUMP'S PATH TO THE NOMINATION - Nate Silver: "The other Republican candidates’ chances of stopping Donald Trump look marginally better today than they did on Friday….Trump remains the front-runner, however, having won 43 percent of the delegates overall among the 20 states and territories to have voted so far.…. If Trump wins [Florida and Ohio], he’ll have a good chance of eventually getting a delegate majority. If he loses both, we might be headed to a contested convention in Cleveland. And if Trump splits them — perhaps the most likely outcome based on where polls stand — we’ll continue to be on knife’s edge….If the Republican nomination were contested under Democratic delegate rules instead, Trump would find it almost impossible to get a majority of delegates, and a floor fight in Cleveland would already be all but inevitable. If every state awarded its delegates winner-take-all, conversely, Trump would be much further ahead, although the bigger swings these rules enable would give his opponents a chance to catch up later on." [538]

Trump could still win nomination even if he loses Ohio - Chris Cillizza: "If Trump sweeps Ohio (66 delegates) and Florida (99 delegates) on March 15, his lead over Cruz would go from less than 100 to nearly 300….Trump would then need to win only 52 percent of the remaining delegates to be the nominee. That's a slightly better pace than the 45 percent of available delegates Trump has won, but he would almost certainly make it to 1,237, with coastal states' votes coming, including delegate-rich ones such as California (June 7) and New York (April 9). If Trump wins Florida but loses Ohio to Kasich— the most likely scenario other than a Trump sweep — Trump's lead over Cruz would be just over 200 delegates. That makes his path to 1,237 more complicated as he would need to win 59 percent of the remaining delegates to formally secure the GOP nomination before the July convention in Cleveland. Should Rubio win Florida and Kasich win Ohio, it's a near-certainty that no candidate will come close to getting the required number of delegates to win the nomination. Trump remains the only one with an even-marginally viable path, but he would need to win 69 percent of the remaining delegates after March 15 if he loses Ohio and Florida. Mathematically possible but practically very difficult." [WashPost]

REPUBLICANS ARE UNDERPERFORMING PREDICTION MODELS - Matthew Atkinson and Darin DeWitt: "After the 2012 election, Vanderbilt political scientist Larry Bartels presented one of the simplest and most robust economic voting models. In this model, the incumbent party’s popular-vote margin depends on just two factors: per capita income growth and fatigue with the incumbent party. He measures those factors using change in real disposable income (RDI) per capita in the second and third quarters of the election year and the number of consecutive terms served by the incumbent party….We replicated Bartels’s model with revised data from the Bureau of Economic Analysis....Basically, the model predicts a dead heat…..with a 51 percent probability that the Democratic Party wins the popular vote….Compare this with the probability in the prediction markets — 65 percent — and you can see that the Republican Party is underperforming this model’s prediction by 14 percentage points….Political scientists have long thought that parties pay a modest penalty, if any, for a contentious nomination process. The current prediction markets appear to believe that the penalty is far more substantial." [WashPost]

The 'PRIMARY MODEL' predicts Republicans will win only with a Trump nomination - Helmut Norpoth: " With Donald Trump as the nominee, Republicans are highly certain to win the presidential election on November 8, 2016. Trump will defeat Hillary Clinton with 87 percent certainty, and Bernie Sanders with 99 percent. These forecasts come from the PRIMARY MODEL. It is a statistical model that relies on presidential primaries and an election cycle as predictors of the vote in the general election….What favors the GOP in 2016 as well, no matter if Trump is the nominee or any other Republican, is a cycle of presidential elections. After two terms of Democrat Barack Obama in the White House the electoral pendulum is poised to swing to the GOP this year….It is possible, of course, that the Republican Party will not nominate Trump as its presidential candidate….The model is 86 percent certain that Hillary Clinton will defeat Marco Rubio. Clinton would get 52.4 percent and Rubio 47.6 percent of the two party vote. According to the model, Ted Cruz or any other candidate (except Trump) would fare the same way against Clinton."[HuffPost]

But could a Trump nomination undermine conventional wisdom in the general election? - Steven Shepard: " If Donald Trump’s coalition carries him to the GOP nomination, it could produce a crazy-quilt general election that scrambles the conventional wisdom about the battleground state map. The Electoral College landscape has been fairly stable over the past four presidential elections: 40 states have voted for the same party in every race since 2000. But Trump’s candidacy — and a unique base of support that’s carried him to victory in places as varied as Alabama and Massachusetts — is raising the prospect of a scrambled November landscape that features longtime Democratic strongholds in play and states that have been firmly Republican for a half-century in jeopardy." [Politico]

REPUBLICANS ARE INCREASINGLY UNHAPPY WITH THEIR PARTY - HuffPollster: "This past October, Republican voters were bearish about their current elected officials and the future of the GOP, but overwhelmingly enthusiastic about their presidential candidates. In that poll, 80 percent pronounced themselves at least satisfied with the Republican primary field, surpassing the percentage of Democrats who said the same.Since then, GOP morale has fallen substantially. While a majority still feel positively about their field of candidates, that percentage is down to 56 percent. Just 41 percent feel positively about the future of the party, and just 30 percent feel satisfied with their current officeholders. Only 11 percent believe the party is more united than it is divided….That doesn't necessarily spell electoral trouble for the GOP….[but Democrats are ]now 15 points more likely than Republican voters to be satisfied with their choices, instead of 14 points less likely to feel that way." [HuffPost]

TUESDAY'S 'OUTLIERS' - Links to the best of news at the intersection of polling, politics and political data:

Ted Cruz is strengthening his hold on second place, according to a new Washington Post/ABC poll. [WashPost]

-NBC/SurveyMonkey's tracking poll finds evidence of Democratic voters aged 25-34 shifting towards Hillary Clinton. [SurveyMonkey]

-Mitt Romney's speech had little overall effect on Donald Trump's support. [Morning Consult]

-Matthew Yglesias doubts Ted Cruz can pose a significant challenge against Trump going forward. [Vox]

-Nate Cohn thinks Cruz's rise compromises Trump's chances of winning a majority of delegates. [NYT]

-David Broockman argues that reforms designed to encourage centrism helped fuel Trump's rise. [WashPost]

-Statisticians get heated over the misuse of the P-value. [538]

-A survey of attorneys finds that one in five lawyers has a drinking problem. [WashPost]