The Ref's battleground index includes Ohio, Florida, North Carolina, Virginia, Iowa, Wisconsin, Nevada, Arizona and New Hampshire. Those states are included because the polls show them to be battlegrounds. Michigan and Pennsylvania are showing signs of closing, but haven't quite gotten there yet. Remember, contrary to conventional wisdom, Trump can win rather easily without Pennsylvania.

With each new poll in the last few days we have seen a major tightening, now to the point where the candidates are undoubtedly statistically tied. Every poll in the battleground states is within the margin of error. One notable exception are the Pennsylvania and North Carolina CBS/YouGov polls, which absurdly assume a 99% turnout, an assumption that bumps Hillary's result by substantial margins. The North Carolina poll was within the margin of error even with the laughable likely voter filter applied.

Today we saw the release of a string of battleground polls that have to frighten the Hillary camp and the media campaign to elect Hillary Clinton. Reuters/Ipsos released the following poll results:

Iowa - Trump 44% Clinton 41%

Maine - Trump 42% Clinton 42%

Michigan - Trump 42% Clinton 42%

New Hampshire - Trump 45% Clinton 44%

Ohio - Trump 46% Clinton 43%

Utah - Trump 35% Clinton 34%

Wisconsin - Trump 39% Clinton 39%

Using the RealClearPolitics averages as guides, Trump leads in Iowa by 0.8 points and Arizona by 2.5 points. Trump trails in Ohio (-3.3), Florida (-2.7), North Carolina (-1.2), Nevada (2.3), Wisconsin (4.0) and Virginia (1.5). A poll in Wisconsin from June was excluded because the data is too old. Two polls in Virginia were excluded because they contain data over three weeks old.

When we apply the latest battleground numbers, we find Trump ahead in Iowa by 1.2 points, trailing in Ohio by 1.75 points and Wisconsin by 2.66 points. When factoring these new numbers, the Ref's battleground index for all battlegrounds shifted from Clinton +2.5 points to Clinton +2.0. The Ref's battleground index in traditionally Republican states changed from Clinton +2.1 to Clinton +1.9.

As an aside, the battleground index does not include states where Trump is leading. The point is to measure how far the person trailing is behind, because that is the measure that actually makes an electoral difference.

The Wall Street Journal ripped Trump in mid-August for running an incompetent campaign. It warned the GOP that "if they can’t get Mr. Trump to change his act by Labor Day, the GOP will have no choice but to write off the nominee as hopeless and focus on salvaging the Senate and House and other down-ballot races." It seems Trump has passed their test. One can imagine that the NeverTrumpers at the WSJ did not expect Trump to actually change his act and move the polls, but he's done just that.

The Ref's Battleground Index

Clinton +2.1 Points in All Ref Battlegrounds (Trump +0.4 from last result)

Clinton +1.9 Points in Traditional GOP States (Trump +0.2 from last result)

What now? NeverTrumpers expected Hillary to hold a commanding lead on Labor Day

Considering that NeverTrumpers want nothing more than to focus only on the Senate and the House and write Trump off, one can imagine that the Wall Street Journal was laying the predicate for doing just that. The NeverTrump theory goes that if Clinton holds a commanding lead, her base will become complacent and opposition to her on the right will mobilize just enough to save Congress. If Trump is close in the polls, as he is, NeverTrumpers argue that fear of a Trump presidency will mobilize Americans to vote in a Democratic congress to offset Trump. If Trump pulls way ahead in the polls, this will cause Americans to almost certainly elect a Democrat House and Senate to oppose him.

First, the NeverTrump theory laid out above concludes that the worst-case scenario is a close race because Republicans are most likely to lose the House in that scenario, in addition to the White House. Nate Silver concludes the contrary, however, that the only way Democrats could win the House and the Senate is a massive Trump loss. Built-in geographic advantages and a decade or more of Republican redistricting virtually guarantee a Republican hold unless Trump loses huge, what NeverTrumpers want. It appears that Silver has the far superior argument based on the facts.

So here we are at Labor Day and it's tied. NeverTrumpers can continue down the irrational path outlined above or the opposite path where support for Trump is much more likely to save the Senate. The House is safe no matter what. We'll see what they conclude this week as this new polls seep into the political psyche.

Please follow my new account on Twitter for daily updates on the election!