The chief architect of the Democrats’ strategy for holding their Senate majority is cautiously optimistic about next week’s outcome. With a heavy emphasis on the cautious part.

In an interview with me, Guy Cecil, the executive director of the Democratic Senatorial Campaign Committee, sounded realistic about the challenges Democrats face, seeming to suggest at one point that keeping the majority might not be the most likely outcome. However, he continued to project confidence, insisting that internal polling had enough races within the margin of error to position Democrats to hold the majority if things go right.

If Democrats do manage to keep the Senate, it will be in no small part due to Cecil’s efforts. He has spent two years putting in place a strategy based on the overriding premise that the only way Democrats can possibly hang on, given the formidable fundamentals they face, is with a serious, sustained, and very expensive effort to change the composition of the midterm electorate.

To that end, Cecil told me that at the end of the day, at least $60 million will have been invested in voter mobilization alone. He also responded to reports that President Obama’s advisers are angry with Democrats for distancing themselves from him. A lightly edited and trimmed down version of our conversation follows.

THE PLUM LINE: In an interview with Bloomberg TV, you seemed to predict Democrats would win Arkansas Alaska, Louisiana, Iowa, North Carolina, and Colorado. The polls have slipped since then. Do you still stand by that?

GUY CECIL: Clearly when you have as many races as we do that are within the margin of error you’re not going to win every race. But our approach has been to keep as many states competitive as possible, and keep it that way until the very end, in the hopes that the right combination of states actually leads to a majority. But no, I’m not predicting wins in every single targeted state. But there’s a path to victory.

TPL: In the polling averages right now, Republicans lead by more than three points in enough states to take the majority, including Arkansas, Alaska and Louisiana. They lead by two-ish in Colorado and Iowa. And they lead in places where you’re hoping to make surprise pickups. Is the polling wrong?

CECIL: There is a history of polling averages getting enough races wrong that it might make a difference when the majority is this close. In Colorado in 2010, 17 of the last 18 polls showed Michael Bennet losing. Our internal polling always showed it a one point race.

For us the focus is more on our own internal modeling and analytics. When we put out a poll from Paul Harstad showing us up one in Colorado, it’s because we believe that is the case. We think with the right formula for what the election looks like in a vote by mail state, with the right percentage Latino turnout and of our vote among Latinos, that we have a margin of error lead that is different from most of the polling averages.

TPL: But the trouble with released internal polls is that nobody knows what the non-released ones say. What is the big picture from all of your internal polling?

CECIL: The aggregate overview is we still have 10 races that are within three points. When we look at Colorado, Iowa, North Carolina, we’re in margin of error races. Statistically these races are even. The bottom line is there are enough races within the margin of error for us to hold the majority.

TPL: A number of these Southern Democrats are clearly under-performing among Southern whites. What are the long term implications if people like Mark Pryor and Mary Landrieu can’t hold on in red states? What does that mean for the Democratic Party?

CECIL: I don’t think they are necessarily under-performing. Most are performing at margins that are above the president’s favorable rating. And most are out-performing what most Democrats would get in those races. We clearly have a challenge in the South. But that’s why we have candidates who have shown an enormous amount of independence.  There’s no mistaking that to be successful we have to maximize our vote among younger voters, unmarried women and African Americans.

TPL: What would it mean if Pryor and Landrieu — who are deeply rooted in their states with powerful brands — can’t hold on?

CECIL: I try not to draw too many conclusions from one election. The country demographically is changing rapidly. Throughout the south, the Hispanic population is growing. In Atlanta, the African American and Hispanic populations are growing. There are significant demographic changes in North Carolina. This is a particularly challenging map. Of the three Senate classes, this is the only one whose states the president lost. The other two classes are in classes that as an aggregate the president won.

TPL: In North Carolina, I’m hearing there’s serious slippage for Senator Kay Hagan among conservative Democrats from the east, which is to say, white voters on the fence who dislike the president.

We haven’t shown any sort of significant decline in North Carolina. We’ve had it at a 1-3 or 1-4 point race. It moves around in there. It’s not a particular area of concern.

TPL: In Colorado and Iowa, the common thread is that Democrats have cast both these Republicans as extremists on women’s health, yet both are at a minimum tied or at a maximum ahead.

CECIL: If you look at the trend in Iowa, it’s since Bruce Braley and his allies have put up the attacks on women’s health that the race started to close. It’s actually a relatively new argument.

TPL: But in Colorado it’s been a longtime thing.

CECIL: The number one indicator [in our numbers] for voting for Mark Udall or against Cory Gardner, is their position on choice. There’s a mis-perception that Mark Udall has only run on choice. He has run ads on flood relief, the government shutdown, the NSA, the economy. But it’s important to note that abortion isn’t just about abortion. It is a reflection of a person’s ideological views and how they view the world.

TPL: Republicans argue that the fact that Udall is still barely tied or a couple of points behind shows that the Michael Bennet playbook [attacks on Personhood and women’s health, which were central in the 2010 Colorado race] has run its course.

CECIL: That’s the same argument Republicans were making in October 29, 2010. Everybody thought Michael Bennet was going to lose. During that period there were op-eds and editorials written claiming he was a one-note candidate.

TPL: The Post reported that some Obama advisers are “exasperated” with Democrats for running away from his record. Couldn’t Democratic candidates have leaned a little harder into making the case that the economy is getting better and that Obamacare is working?

CECIL: The president has done everything we have asked him to do. But this is not about the president. It is about the map. It is our job to make Senate races about the two people on the ballot. Our advice to candidates is that when somebody disagrees with the president, they should say so, and that when somebody agrees with the president, they should say so…For us to nationalize the election in a series of red states would play into the terrain that Republicans want us to play on.

TPL: But the Republican strategy has been about layering Ebola on top of ISIS and the child migrant crisis to nationalize these races and make them not about a choice between two candidates. Hasn’t the President’s handling of some of these things made it easier for Republicans to do that?

CECIL: The Republicans’ goal is to nationalize the election. Whether they were going to use Obamacare or Ebola or something else, this is the strategy they have laid out. Our strategy is not counter to the president. It’s making the race about education in North Carolina. In Georgia, it’s talking about Michelle Nunn’s long record of encouraging people to engage in volunteer service and comparing that to David Perdue’s record of outsourcing jobs.

Our job for the next six days is to win this election. It is not to figure out whether somebody should have been more or less involved. I believe that if the DSCC and our candidates and our allies stay focused on that, we have a shot. Despite the map, despite turnout, despite the midterms, we are in a position — doesn’t mean it’s guaranteed; doesn’t even mean it’s the most likely — to hold the majority.

TPL: At the end of the day, how much will have gone into field and voter mobilization?

We will end up somewhere between $60 and $65 million in our targeted states that is either transfered or raised by the DSCC or our candidates. Last week we knocked on 143,000 doors in Colorado. We have registered tens of thousands of voters in a number of states.

Turnout is required but not sufficient. We must engage in trying to change the electorate. But this is not just a turnout election. If it were just turnout election we would always win these tough states. But we still have persuadable voters out there that we have to also stay focused on. So [turnout] is not the only thing that’s going to lead to a victory. But that investment hopefully will pay off.

***************************************************

UPDATE: Post edited slightly for clarity.