A new CBS/YouGov poll purports to show Nimrata ‘Nikki’ Haley gaining on President Donald Trump in New Hampshire. The top lines show Haley with 29 percent, up from just 11 percent in the equivalent September data. There’s a few things that stand out about this data, upon first glance:

1. The September poll quizzed 943 people with a margin of error of 3.7 percent. The new, December poll quizzed 855 voters with a margin of error of 4.1 percent. But that number drops dramatically to just 459 on the question of which candidate respondents would vote for. No such alteration of sample size occurs in the September poll. This also means the margin of error changes, though CBS and YouGov don’t report that. The Hill reports the latter actually has a margin of error of 5.5 percent;
2. The question of presidential candidate support also appears to oversample independents, commingling the results with Republican voters to give the impression that Haley is gaining massive support. September had 43 percent of independents polled, whereas December sees that number jump to 45, leading to a decrease in self-identified Republicans polled;
3. Amongst Republicans only, Trump score 52 percent, with Haley on 24 percent, and DeSantis on 17 percent. Amongst independents, Trump score 41 percent, and Haley gets 34 percent. This explains why Haley and her allies are openly recruiting non-Republicans to vote on January 23rd;
4. The data further suggests Haley is mostly gaining with those in the age bracket 45-64, more than doubling her vote amongst conservative voters, but almost tripling her vote amongst independents;
5. Confirmation of how the data skews social liberal/independent comes in other significant shifts. Examples:
   1. In the September poll, 15 percent said they want abortion to be “legal in all cases”. This rose to 19 percent by December: a four point, but 26 percent shift;
   2. In September, 41 percent said they “don’t like Trump personally”, with this number jumping 25 percent and 10 points to 51 percent in December.

None of this is to say that Haley’s hasn’t legitimately increased support in the Granite State. It would logically follow after the endorsement of the popular state governor Chris Sununu. But not only does the gain appear exaggerated by the different sample used, it also confirms a number of weak spots for Haley:

1. Haley’s top characteristics appear to be her “likability” and “reasonableness”. These are easily assailed traits, and she notably drops off on the policy end, polling below 5th place Governor DeSantis for “leadership” and conservative values;
2. 80 percent of respondents say they want a President who will “deport millions of immigrants they believe to be undocumented”. Haley is a noteworthy squish on the subject, recently telling an audience she believes corporate America should set the country’s immigration priorities;
3. Trump outperforms Haley on the critical metrics: strong leadership, preparedness, and conservative credentials.

Is Haley gaining some support in New Hampshire? Sure. But her ceiling is low and her support is weak. She hasn’t really taken anything from Trump, instead picking up a couple of percentage points here and there from DeSantis, Vivek, and others. Where she really gains is when you include left-leaning independents in the data. That’s why she wants those people voting in the New Hampshire primary. The same as Casey DeSantis wants in Iowa. A stitch-up.