2016 Postmortem

some white male voters are a subset of voters who switched. Your logic is flawed. Of course, it is possible to approximate how many voters did the switch by a combination of corrected exit polls and actual vote numbers per location. Breakdown by ethnicity is possible as well. Because it is statistical it will not be 100% accurate, but the data will have a confidence margin strong enough to draw conclusions. Your post has no data, just statements of "fact" without support, and therefore is itself completely unsupported.