Higher model complexity always leads to lower bias and lower variance.

As model complexity increases, bias typically decreases, but variance typically increases, potentially leading to overfitting.

Bias and variance are independent of model complexity.

Lowering the variance requires increasing the bias to zero.