South Carolina Senate: Graham Holds Wide Lead

A new Abacus Data survey gives Republican Darline Graham a 13-point edge over Democrat Annie Andrews, 55 percent to 42 percent, in the South Carolina Senate race. The poll, completed August 29 with a sample of 2,500 likely voters, tracks with PollingSource's Likely R rating for the contest and offers little evidence that Andrews has narrowed the gap since earlier in the cycle.

The size of the sample is notable on its own — 2,500 respondents is a larger-than-typical universe for a statewide poll, which should tighten the margin of error and lend the topline numbers somewhat more weight than a standard 800-voter survey. That does not immunize the result from the usual caveats about partisan lean in the sample or turnout modeling assumptions, but it does mean Abacus is not working with a thin reed. A 13-point spread built on a 2,500-person sample is a harder result to dismiss as noise than a similar margin drawn from a smaller poll.

What stands out is the divergence from South Carolina's other marquee statewide race this cycle, the governor's contest, where recent polling has shown a narrower margin. If that pattern holds, it would suggest South Carolina voters are not simply defaulting to a single partisan lever up and down the ballot — Graham may be outperforming the top of her party's ticket, or Andrews may be underperforming a generic Democrat in a state where Republicans hold a durable structural advantage. Either explanation matters for how strategists in both parties allocate late-cycle resources. A campaign that assumes uniform swing across a state's statewide races can misjudge where money and turnout operations are actually needed.

It's also worth treating the comparison with some caution. Different pollsters, different fieldwork windows, and different likely-voter screens can produce diverging margins even when underlying voter sentiment hasn't shifted much. Without a single pollster tracking both races on the same sample and timeline, the "divergence" could partly reflect methodology rather than genuine ticket-splitting. Still, the gap is large enough — and the Abacus sample large enough — that it's not easily waved away as pure survey artifact.

For Andrews, the math is not encouraging with roughly two months remaining before voters go to the polls: a 13-point deficit in a state carrying a Likely R designation leaves a narrow runway. For Graham, the number reinforces her position as the clear favorite, though a 55 percent topline is not so dominant as to close off scrutiny entirely. The next test will be whether subsequent polling, from Abacus or other pollsters, confirms this margin or reveals it as an outlier — and whether the governor's race narrows further or the Senate contest tightens to match it.

Worth watching in the weeks ahead: whether the two statewide races continue to move independently, or whether one eventually pulls the other toward it.