Michigan · 2016 President

The column below is not a finding that interference occurred, and it does not prove that any candidate actually received extra votes. It is a rough advisory screen from the loaded review indicators. For down-ballot comparison flags, the app compares same-party presidential votes against a comparison race such as U.S. Senate or Governor, then labels the direction with the larger relative gap. Governor-only comparisons are marked low confidence because they can mostly reflect ordinary race-specific ticket splitting. When the comparison is U.S. House, the column names presidential-over-House dropoff direction instead of candidate benefit because House races are district- and candidate-specific controls.

That direction can be affected by split-ticket voting, incumbency, local candidate strength, undervotes, uncontested races, one-sided comparison races, reporting-unit definitions, missing comparison rows, or ordinary political geography. Treat it as "if this pattern needs review, this is the side the math points toward," not as a causal claim or a conclusion.