State Senator Jeff Van Drew (D-1) is a “strong frontrunner to win the open seat in New Jersey’s Second Congressional District and return the seat to Democrats’ hands for the first time in 24 years,” according to a TargetSmart polling memo.

In the poll of 500 likely mid-term election voters conducted by the firm from April 19-24, 2018, Drew comfortably leads the presumptive Republican nominee, Hirsh Singh, 52% to 34% in an initial head-to-head question for the seat occupied by retiring U.S. Rep. Frank LoBiondo (R-2).

Van Drew holds a large advantage despite the Republican leanings of the district, the survey finds. In the same survey, a generic, unnamed Democrat trails a generic, unnamed Republican candidate by 7 points, 39 to 46 percent. Self-identified Republicans outnumber self-identified Democrats in the poll by a double-digit margin (41 percent identify as Democrat or lean Democrat and 51 percent identify as Republican or lean Republican). Despite those factors, the well-known Van Drew starts with an 18 point lead, and he is known by more than 60% of voters.

“As we expected, Senator Van Drew’s immense popularity and cross-party appeal in the Republican-leaning First Legislative District propels him to a strong lead over his opponent. He strongly over-performs what the raw partisan trends in this district would predict, and is in an enviable position for a Democrat running for an open Republican seat,” said Ben Lazarus of TargetSmart.

Van Drew’s poll numbers come on the heels of many of the top political forecasters changing their ratings of the race to “Leans Democrat.” Cook Political Report, Larry Sabato’s Crystal Ball, and Real Clear Politics have all moved NJ-02 from “Toss-up” to “Leans Democrat” since Van Drew entered the race.

The poll, according to the memo, reached a representative sample of 500 voters in NJ-02, and was weighted to be representative of likely midterm election voters. Two thirds of respondents were reached on landline phones and one third were reached on cell phones. The margin of error is +/- 4.4%.