It’s been three weeks since the last CBS/YouGov poll, and an eventful three weeks at that. During that period of time, the media and Democrats have covered almost nothing except the family-separation issue in border enforcement, with both having at times to admit that the same outcomes took place in the Obama administration, although not on the same scale. For the past fortnight, the American electorate has had a ringside seat to this overwhelming firestorm. With the midterms just four months out, the political impact of this has to have been massive. Right?

Maybe not. Three weeks ago, the previous CBS/YouGov poll showed Democrats leading Republicans by five. This weekend, their latest poll shows Democrats leading by … four?

The question was asked differently in the previous poll — “If the 2018 election for U.S. House of Representatives were held today, whom would you vote for in the district where you live?” That’s not too much different, however, and the latest poll question more precisely mirrors the generic-ballot questions in other polls. With a margin of error of ±2.6%, not only is their no significant difference, the latest poll results comes close to a dead heat in the generic ballot.

Nor is that the only curious result. The Reuters/Ipsos poll series shows more of a change than the CBS/YouGov series, but in the direction of the GOP. In polling done in the second week of June, Democrats were up 10 points; last week’s tracking results show Democrats only up six. The Economist/YouGov polling listed at RCP shows Democrats gaining only a single point in the same time frame. Quinnipiac has the Democratic advantage declining by a point since the end of May to last week. Only CNN, which polls every six weeks, shows a significant gain (+5) for Democrats, and that takes into account all of May as well as the first three weeks of June.

The rest of the numbers on immigration in this CBS/YouGov survey are not good news at all for Republicans facing tough midterm elections. CBS covered those in detail over the weekend, but added a significant caveat:

One-third of Americans and registered voters say the separation policy will be a very important matter in their congressional vote this fall. Of those who say it will change, more say it will make them vote Democratic than Republican. That gap is even larger among those who have never voted in a midterm before, meaning that if Democrats are to try to use discontent over this issue, it might be necessary for them to get new voters to the polls.

How successful will they be in doing that? The results from the generic ballot suggest that they’re not succeeding so far. Democrats may boost turnout in their core areas of support, but that’s not where they need to convert House seats. To get that kind of win, they need the type of wave they produced in 2006. At that time, they led generic ballot polling by double digits across a number of series; right now, their RCP average is +6.0%, and it’s not going up so far during the immigration debate. And after the perfect storm that Democrats got the past two weeks on this issue, it’s tough to see how they can expect those numbers to move significantly from the status-quo position that gap suggests.