Except most references are complete BS anyway. Normally good references are agreed upon before leave, regardless to if they were good employees or not.
Bad references are normally given because of political reasons(as good references are almost always given) not performance. I.E Employee exposed bad practices or made boss look bad, and a bad reference is used as retaliation.
I think there was some academic research that said references are the worst predictors for performance.
This is the problem we're trying to fix with my startup socialcheck.me. We're trying to make a safe environment to give full picture comprehensive reference feedback by ensuring privacy and anonymity after verifying relationships are real via social networks. Then we make the data easy to analyze and compare. But admittedly, it's been difficult to figure out even then how to get people to trust that we keep the feedback private and anonymous the way kids trust Snapchat to really delete messages. Good reference checks are hard and broken (which is why we're trying to fix them).
Doesn't fix the problem of references being used for politcal retaliation, rather than actual performance. Which what the large chunk of bad references are about. In-fact it increases it, because the employee can no longer sue for retaliation, so the manager can make up anything he wants. Managers using references to retaliate against employees is fairly common.
This is why people are scared to whistle-blow. And businesses who value integrity miss out on good employees.
Yeah, we try to compensate for that by increasing the data volume to balance out outlier feedback and identify whether something is a trend or not. But it's not easy.
Bad references are normally given because of political reasons(as good references are almost always given) not performance. I.E Employee exposed bad practices or made boss look bad, and a bad reference is used as retaliation.
I think there was some academic research that said references are the worst predictors for performance.