Regressing human rights on democracy is definitionally circular, not causal testing.

Evidence

It is not uncommon to find human rights as a dependent variable where democracy is one of the independent variables. By definition, one cannot be a high-quality democracy and extensively violate human rights: respect for human rights is built into the conceptualization of democracy (particularly following the Ideal Type Guideline). Another example is the literature on flawed elections as a cause of democratization (e.g., Morgenbesser and Pepinksy 2019), where the problem of tautology is explicitly recognized. — Gary Goertz, Social Science Concepts and Measurement, 542

Categorization

• SHARE: A pointed tidbit for anyone in comparative politics or human rights research who treats democracy indices as exogenous predictors.
• SEED: A methods piece: ‘Your Independent Variable Is Your Dependent Variable’ — cataloguing famous democracy-and-rights regressions that can’t fail to find significance.