96-hours till votes are counted in the election for US House of Representatives from California's 52nd Congressional District, incumbent Democrat Scott Peters and Republican challenger Carl DeMaio are today as they ever were: locked in hand-to-hand combat to wrest control of the seat. Today, SurveyUSA shows it: DeMaio 45%, Peters 44%. The research was commissioned by KGTV-TV 10News and the UT San Diego newspaper.

This is the 4th pre-election tracking poll that SurveyUSA has conducted in CA-52. DeMaio has led, outside of the theoretical margin of sampling error, only once, the first time, back on 06/13/14. Since then, the nominal lead has seesawed, but always within the theoretical margin. Today, heading into the final weekend before votes are counted, and with many CA ballots already returned, both candidates are well-positioned for victory; neither candidate can be considered the favorite.

If Peters holds the seat, which he may, it will be because young voters turned out in larger numbers than here forecast. And it will be because Latino voters voted in larger numbers than here forecast, and may also have voted more Democratically than here shown. If DeMaio wins by a narrow margin, it will be because seniors turned out in larger numbers than here forecast, and because white voters represented a larger portion of the electorate than here shown. If DeMaio wins by more than a couple points, that would be evidence of a much talked-about "Republican wave," which may affect local elections across the country, as part of which a number of sitting House Democrats may get washed away.

Consistent in all 4 of SurveyUSA's pre-election polls, voters say that "integrity" is the most important issue in the contest. And on the issue of integrity, Peters leads DeMaio 48% to 40%. Fiscal responsibility is the next most important issue, and on that issue DeMaio leads decisively, 74% to 21%.

Atypically, the number of undecided voters is up, as Election Day approaches. It has doubled since June. This is sometimes an indication that each candidate has succeeded in raising doubts about the other. And it is sometimes an indication that voters are holding their nose while marking their ballot.

This research was conducted 100% by telephone; live interviewers were used to hand-dial cell phones: SurveyUSA interviewed 580 registered voters from CA-52 10/27/14 through 10/30/14, using Registration Based Sample (RBS, also known as Voter List sample), purchased from Aristotle in Washington DC. Of the registered voters, 551 were identified by SurveyUSA as having already voted, or as being likely to vote on or before the 11/04/14 General Election deadline. This research was conducted 100% by telephone. Respondents reachable on a home telephone (68% of likely voters) were interviewed on their home telephone in the recorded voice of a professional announcer. Respondents not reachable on a home telephone (32% of likely voters) were interviewed by live operators, who hand-dialed cell-phones, secured the cooperation of the respondent, qualified the respondent, conducted the interview, logged the answers, and remained on the call until its conclusion.