Democrats should be very happy with yesterday's new R2K poll Mississippi Senate poll showing Ronnie Musgrove losing to interim Sen. Roger Wicker by a 48-to-43 split. The last R2K poll from late July had Wicker leading only 45-to-44, so Wicker has clearly put a bit more distance between him and the former governor.
The last two polls of the race, both from Rasmussen, showed Wicker ahead by nine points after having the race virtually tied. The Ramussen polls had many people questioning, this author included, whether Musgrove's chance had passed. (Unfortunately, you have purchase Rasmussen's cross-tabs, so we can't look at any of them here.) After Palin was announced, and McCain's numbers exploded in the South, I believed Musgrove would probably be down 15 points right now. These results question the impact of the balance bounce on individual down-ballot races, even those in the Deep South.
They key, however, is this: while Musgrove is down five, the numbers below the surface are very favorable to the Democrat. While Wicker's base, the white vote, is nearly all decided -- with just four percent undecided, Musgrove base, the black vote, is 18 percent undecided. In other words, Musgrove has appreciably more room to grow than Wicker, and if he can rally those undecided black voters into his camp in the same numbers as the decided black voters, while holding onto his nearly-one quarter of whites, he will win.
To prove this point, let's look more closely at the cross-tabs, which Kos always provides.
Musgrove's ratings have surprisingly risen since May
In yesterday's poll, Musgrove's favorable/unfavorable split stood at 50-43 (it was 48-41 in July, and 46-33 in May), while Wicker's registered at 52-41 (51-39 in July, and 52-25 in May). Thus, we see that both men had a slight rise in both of their numbers: almost identical between the two men. What this probably means is that people who like the two men have not changed their views.
What is even more interesting is that Musgrove's favorables have actually risen four points between May and today. This is amazing because of the two candidates, Musgrove is much more well known in Mississippi, as he has been elected to two statewide offices while Wicker has only been in the Senate for less than a year after being a congressman for ten years. Therefore, one would expect Musgrove's favorable numbers to be more solidly entrenched than Wicker's. These numbers show that this is not so.
Looking at the unfavorables is also instructive. In May, when Wicker had only been in office for a few months, he had a 52/25 personal split, meaning that he had a solid approval in lieu of his party affiliation, but his negative was low because he was unknown. However, the polls show that his unfavorables have jumped from 25 to 39 to now 41. Apparently, Democrats' attacks on Wicker have been successful not so much in impacting his positives, but in jacking up his unfavorable number 16 percent. Conversely, Musgrove's unfavorable number has gone from 33 to 41 to 43, an aggregate bump of 10 points. What I would attribute this to is the GOP's attacks, and reminders of what they argue were Musgrove's various failures as governor. Still, Democrats' attacks have been slightly more successful if for no other reason than the fact that Wicker's negatives have jumped more than Musgrove's.
The key numbers among whites and blacks
The full breakdown is even more valuable. Among men, Wicker leads 52-40 (49-41 in July, 50-39 in May), while Musgrove has a slight edge with women, 46-44 (47-41 in July, 45-42 in May). Among Democrats (40 percent of the sample), Musgrove wins 84-7, while Wicker wins GOPers 86-6, and indies are virtually tied at 46-45 Wicker (46-43 Musgrove in July, but 45-43 for Wicker in May).
This information makes more sense when you look at the race breakdown. Wicker wins whites 73-23 (67-26 in July, 65-26 in May), while Musgrove wins blacks 75-7 (75-8 in July, 73-9 in May).
These numbers tell us a few things. First, despite all of the negative attention being focused on Musgrove regarding his connections to some figures in the federal trial involving that failed Mississippi beef-packing facility, his favorable ratings among individual groups are much the same as we noted above. Clearly, after being well-known in Mississippi for years, most people are pretty hard in their view of him, one way or the other.
Second, Musgrove lost three percent of whites, while Wicker gained six percent. A good part of this may be the Palin selection, and perhaps even the negative ads and stories that have been launched against Musgrove. Furthermore, I think it was plainly inevitable that Wicker would ultimately end up with at least three-fourths of whites.
Still, at 23 percent, Musgrove is doing fairly well among the conservative white population for a Democrat running statewide. If he can hold onto that total, he is halfway home. It is important to note that there are only 4 percent undecided, providing some evidence that that portion of the electorate is more hardened in its support of the two candidates (and it stands in contrast to the black electorate, as will be seen).
Third, and most importantly, these results pretty clearly show us precisely what Musgrove needs to do to win. His support among the black electorate, at least right now, is not where it has to be for the former governor to pull of the upset. It is fascinating that the black vote has not really changed between polls, as Musgrove's number stayed at 75 percent, and Wicker's dropped just one point, to 7 seven percent (it was 75-8 in July, and 73-9 in May). A whooping 18 percent of Mississippi's blacks are still undecided right now.
As we have discussed in our past posts, Mississippi has the largest percentage of black residents of any state, coming in at around 37 percent of the Magnolia State's population. In statewide elections, Democratic candidates routinely win 90 percent of the black vote, while Republicans take 80 percent or more of the white vote en route to victory. For a Democrat to compete in Mississippi, he or she has to win nearly all of the black vote, and somehow crack 20 percent of the white vote.
As the latest R2K poll shows, Musgrove is around where he needs to be with whites. Again, while 23 percent sounds pitiful, in a state as conservative as Mississippi, it is not. Where Musgrove is lagging behind is with the black electorate. Perhaps this is not terribly surprising. In the past, Musgrove has been criticized for some of his initiatives and actions as governor, and for not doing more to reach out more to the black community in the state. Rep. Bennie Thompson of the Second District has been particularly critical of Musgrove in the past.
While it is not clear that these same issues are the cause for the large number of undecided voters, Musgrove must make sure wins nearly all of the undecideds to have any chance of winning. As we will now discuss, it would also help if black turnout exploded as some commentators have suggested will happen with Barack Obama at the top of the ballot. In fact, for Musgrove to win, that may have to happen.
Let's take a look at some turnout models, and explore what percentages Musgrove will need to achieve in order to win.
Black turnout models
Model #1: Black turnout is the same as it was in 2004
In the 2004 presidential contest, exit polling showed that the vote was 34 percent black and 65 percent white. If black turnout were to remain at this level in November, even if Musgrove got 25 percent of the white vote -- clearly not a given -- he would need over 99 percent of the black vote. This is not going to happen. Thankfully for Musgrove, there is no way black turnout is going to be this low.
Model #2: Turnout is equal to the state's composition
This scenario is certainly possible, though I do not think very likely either. Assuming that the vote is 63 percent white and 37 black, mirroring the state's own composition, there are various ways Musgrove could pull out a win.
If he got 25 percent of the white vote, he would need 92.6 percent of the black vote. This is doable, but it would be tough. If it ticks up to 24 percent -- and if the latest poll is accurate, this is possible; with 4 percent undecided, it is fair to say that could break 3-to-1 to Wicker, making the final breakdown 76-24 -- Musgrove would then need about 94.3 percent of the black vote. Still a steep hill to climb, but not impossible.
What if Musgrove stays at 23 percent? He would then need around 96.27 percent of the black vote. This is very unlikely, in my opinion.
Model #3: Black turnout increases
While you may accuse me of being overly-optimistic, I think it is likely that black turnout will increase past 37 percent. The only really question is how much? Forgive me for breaking into so many numbers, but the movement of even one percent, as will be seen, could be critical to Musgrove's chances
Let's say Mississippi gets 38 percent black turnout:
If Musgrove gets 23 percent of the white vote, he would need 94 percent of the black vote;
if 24 percent, he would then need 92.45 percent of the black vote;
if 25 percent, he would then need 90.8 percent of the black vote;
This latter number is very doable for Musgrove. It would be about the same percentage John Kerry got in 2004, and it is a number most Democrats get in statewide votes.
Let's assume Mississippi gets 39 percent black turnout:
With 23 percent of the white vote, Musgrove would need 92.26 percent of the black;
if 24 percent, he would need 90.7 percent;
if 25 percent, he would need 89.1 percent;
As is obvious, Musgrove's chances get better as black turnout increases.
Now let's say black turnout hits 40 percent:
With 23 percent of the white vote, Musgrove would need 90.5 of the black vote.
if 24 percent, he would need 89 percent;
if 25 percent, he would need 87.5 percent of the black vote.
If 41 percent?
With 21 percent of white, Musgrove would need 91.7 percent of the black vote;
if 22 percent, he would need 90.3 percent of the black vote;
if 23 percent, he would need 88.9 percent of the black vote;
if 24 percent, he would need 87.5 percent of the black vote; and
if 25 percent, he would need 86 percent of the black vote.
For fun, what if black turnout hit 42 percent?
With 21 percent of the white vote Musgrove would need 90.1 percent of the black vote;
if 23 percent, he would then need 87.3 percent;
if 24 percent, he would need 85.9 percent of the black vote.
if 25 percent, he would need 84.5 percent.
It is pretty darn clear that is the black vote can somehow hit or clear 40 percent of the total vote, Musgrove would actually have a spectacular shot to win.
Analysis
Of course a big caveat is that these numbers are all estimates, and the lack of exact percentages probably skews the margins a bit; R2K probably rounded figures up and down. But since we do not possess the exact data, we have to make do with what we do have. Furthermore, as the 2004 exit polls showed, there is a tiny percentage -- definitely no more than two percent, and likely one percent or less -- of the state's voting population that is not white or black, and we have not included that small subset in our models. My guess is that this tiny bloc would disproportionately favor Democrats more than Republicans, and thus they could provide a small bump to Musgrove.
So what are our conclusions from this data? Pretty simple: the larger the black turnout, the better it is for Musgrove. I think it is very fair to say that it will surpass 37 percent, but it is impossible to predict by how much. Maybe it will hit just 38 percent, or maybe it could go to 42 percent. Quite honestly, I can't begin to guess with any certainty.
Naturally, Musgrove is going to need to work extra hard with the black community. If Rep. Thompson's criticisms are accurate, Musgrove could have some lingering problems in the community. The fact that such a large group of black voters has remained undecided -- 18 percent undecided this week, 17 percent undecided in July, and 18 percent undecided in May is certainly concerning. Though, the breakdown among blacks over the three months -- 75-7 in September, 75-8 in July, and 73-9 in May -- shows that Musgrove has pretty held a steady three-fourths plus of the population in the state, while Wicker's support among blacks has been stagnant, and remained below 10 percent.
Musgrove can also take some solace in the fact that there is surprisingly a fairly large number of undecided voters in the presidential contest.
September 11: Obama leads 80-5 with 13 percent undecided;
July: 81-4 with 15 percent undecided; and
May: 79-5 with 16 percent undecided.
Indeed, while there have always been slightly more undecided black voters in the Senate contest, the fact that there is still 13 percent of black undecided for President is very interesting. Frankly, I don't believe the number, particularly for the Obama race. Obama is going to win the black vote around 95-5 in the end, if not by more. I do not think there is any question about that.
Musgrove's success among black voters will not match Obama's. Duh. However, his number should not fall below 90 percent, despite the high undecided number. Let's look more closely at this.
Assuming the percentages from September hold, and undecided black voters break along the same 75-7 margin that decided voters have, Musgrove should get 91 to 92 percent of the black vote, with Wicker getting between 8 and 9 percent.
Of course, if Musgrove really does have big problems with some black voters, then maybe this remaining 18 percent undecided may not break along to Musgrove in as great numbers as they other 83 percent already have. This is entirely possible.
Let's then say for the sake of argument that this remaining 18 percent breaks 85-15, instead of 75-7. In that case, the margins really would not change much, and Musgrove would still end up with around 90 or 91 percent.
What if Wicker gets a quarter of the remaining undecided blacks? Then the breakdown would be around 88-12 or 89-11. It is hard to envision Wicker getting one-fourth of the remaining black voters -- that is just such a huge number for a Republican candidate -- but if he did, it would make a huge difference, and I am sure this is a group Wicker has been trying to reach. Assuming average or even slightly higher black turnout (in the range of 38-40 percent), that up-tick for Wicker would require Musgrove to get 25-26 percent of the white vote to cobble together a bare majority. (In terms of third parties, the two men are the only names on the ballot, which makes calculating these things easier.)
All of this is why this election could be determined by just a tiny move in percentages. If Musgrove gets 24 as opposed to 23 percent of the white vote, or 92 as opposed to 89 percent of the black, that could make all the difference in the world.
Applying the fresh polling data to predict an ultimate result
Assuming that the latest poll is accurate, whites are currently breaking 73-23 (with four percent undecided) to Wicker, and blacks 75-7 (with 18 percent undecided) to Musgrove, and turnout mirrors the state's composition -- a big if -- then the end result would be Wicker winning the white vote 76-24, and Musgrove winning the white vote 91.5-8.5, or approximately 51-to-49 overall, a nail-biting win for the Republican.
Should these percentages hold, but black turnout increases to 38 percent, Wicker would still win by an even closer 50.35-to-49.65,
If black turnout were to hit 39 percent with the polling data holding, Musgrove would overtake Wicker and win by 50.325-to-49.675. We have critical mass!
And just for fun, if black turnout hit 40 percent with the existing poll data from R2K, Musgrove would win by a slightly wider 50.775-to-49.225.
The lesson? Simply that Musgrove has a clear path to victory if black turnout can hit 39 percent, and he can hold onto what he has now in terms of 23 percent of the white vote and 91.5 percent of the black vote, while at the same time hoping of course that the undecideds break along the same percentages as the rest of the voters (which may be more questionable).
Do you think Governor Barbour hasn't looked at these numbers? If Musgrove gets 24 percent of the white vote in November, essentially what he has now, and the 18 percent undecided among the black electorate break 91.5-to-8.5 to Musgrove, all he needs is for black turnout to increase by just two percent over the state's overall black population. This is entirely doable, and it is likely that this scenario has been giving huge heartburn to the governor.
Barbour ballot move therefore makes political sense
Processing these numbers make it very easy to understand Governor Haley Barbour's recent power-play to move the special election to the very bottom of the November ballot. Barbour isn't stupid. In fact, in all honesty he is a pretty brilliant (albeit unscrupulous) political strategist. Take another look at our simplistic turnout models. If black turnout exceeds 37 percent, Musgrove's mission becomes so much easier. Say black voters turnout at 40 percent -- a distinct possibility -- and Musgrove wins an even 23 percent of whites. He would then need 90-91 percent of the black vote, and he would win. This has to be on Barbour's mind.
Furthermore, that it is a special election provides one more small advantage to Musgrove. Under a quirk in state law, special election candidates are given no party designation on the ballot. As a result, both Musgrove and Wicker will appear without an "(R)" or "(D)" next to their names. Without a doubt, I think that with tens of thousands of commercials, ads, campaign appearances, and mail pieces filtering across the state, just about every person in Mississippi will know who is who, so the advantage Musgrove might gain from this is very minimal. That being said, in a state as conservative as Mississippi, if even one percent of voters went to Musgrove in the mistaken belief that he was not the Democratic nominee, that one percent could prove critical.
All of this why Barbour first moved the date of the election to November, and then moved this race to the bottom of the ballot. He has no idea what how many black citizens will vote in November. Nor does he know if Wicker can break double-digits with the black vote. He may be looking at very close internal polling from the Wicker camp and the NRSC which only exacerbates his fears, but the fact is that black turnout is unknowable, even with good polling. If black turnout were to hit 40 percent or more, every thing else could fly out the window, and in that case, Musgrove (assuming that 18 percent of blacks undecided breaks to him as it should for a Democrat) would have a clear path to victory.
The white vote too has to be concerning to the Republicans given the available data. In May, Wicker led among whites 65-26, then 67-26 in July, and most recently 73-23. This data reveals a couple of salient points. First, Wicker has consolidated the white vote. This was inevitable: no Republican, even a crummy one is going to get less than 70 percent of the statewide white vote. Second, the litany of Republican attacks on his gubernatorial tenure and stories trying to tie Musgrove to the beef plant case have clearly moved some votes away from the Democrat.
The question is whether Republicans can move any more of those voters out of his camp. You can rest assured that Republicans will continue their attacks on Musgrove to try to get his number down to 20 percent. My sense is that is certainly possible, but it won't be easy given that Musgrove spent eight years as Lieutenant Governor and then Governor, so clearly he has the longtime support of a portion of the white population, however insignificant it may seem. Plus, with the undecided white vote dropping from 9-to-7-to-4 percent shows that white voters' are coalescing around a candidate, and moving them later may be a bit tricky.
The bigger matter for Republicans may be the black vote. Barbour's calculation is that clearly, black turnout will increase by some amount. Given that it was at 34 percent in 2004, there will be many new voters and other voters who do not often cast ballots in elections, even national ones. Moving the race to the very bottom of the ballot, simply by common sense and experience, will dilute the numbers of voters casting a ballot in the race. This phenomenon is not atypical anywhere, whether in Mississippi, Hawaii, or New Jersey. Therefore, Barbour's objective here is to create some drop-off because he believes that the new voters are more likely to be Obama, and by extension Musgrove voters.
Conclusion
In sum, there are two factors that will determine this election: first, whether Musgrove can hold 23-25 percent of the white vote, and second, how high black turnout will be and whether the 18 percent of undecided black voters will still break for Musgrove in the end. Republicans can try to impact the first one, but the second is largely unknown, which is why we have the ballot trick.
One thing that Musgrove certainly has in his favor is Barack Obama. R2K's poll also found the presidential race at McCain 52, Obama 39, after being 51-42 in July, and 54-39 in May. For the most part the numbers have been pretty stable, but the key number is that Obama's deficit has an average of around 12 percent. No one would argue that Obama has any chance of winning the state outright, but if Obama can end up losing by 12-14 percent, that would show that black turnout is enormous (Kerry lost Mississippi by 19 percent). In that case, those are voters that, according to our models, will favor Musgrove.
My own gut read on the race is this. I just can't see a Democrat polling poorly among blacks statewide, particularly with Obama on the ballot. I fully expect Musgrove to win the black vote around 90-10. The big questions then are (1) can he hold onto 23-24 percent of whites; and (2) just how many blacks will go to the polls on election day? Barring any really negative stuff coming out about him, I think Musgrove can hold onto that near-quarter, leaving the outcome up to black turnout. It is also a big question how much the ballot issue will impact that number.
Musgrove certainly still has an uphill battle to win this November, but if the R2K polls have been accurate, he absolutely has a path to victory. The white vote, Roger Wicker's base, is nearly all decided and less fluid than the black vote, which is 18 percent decided. Therefore, Musgrove has much more room to grow. This is particularly evidence when we examine the data above which shows that while Musgrove had 26 percent of the white vote in May, the number has been fairly stable for the last four months. Furthermore, as we found above, if the undecided voters break along the same lines as the decideds, and black turnout goes up to 39 percent, he will squeeze out the barest of victories.
It is clearly going to be a tall order, but no matter who wins, the race is going to be very close.
Links: May poll, July poll, and September poll.
Note: A state circuit court judge today ruled that the special election must move to the top of the ballot, just below the presidential race. This is a win for Musgrove, but given the composition of the state supreme court, it would be a surprise, at least to me, if this decision is ultimately upheld by the Mississippi high court. Stay tuned.
Posts
1 comment:
Mark, you contemplating any posts on potential electoral impacts of the 2010 census? Might make for a good post. A fear Dems must have is in the losses that will be felt by blue and blue-purple states.
Post a Comment