To be fair, these studies leave a lot to be desired. It is simply a fact that cheaper cars will tend to score more highly than expensive cars because not only is there less to go wrong, the tolerances can be looser, so they need repairs less often and when they do need repairs, they will cost less.
For example, an expensive car's interior panels may be manufactured to a template size ± 0.5mm. In contrast, a cheap car's panels may only be manufactured to ± 2mm, so they cost less to make.
Where this starts to hurt 'premium' brands in the reliability stakes is that because expensive cars are generally less reliable and more expensive to fix than cheap cars, and because premium brands have a disproportionate number of expensive cars, they will tend to score lower in these surveys than mainstream brands, even though class for class (e.g. Porsche Boxter vs Nissan 370z*), the reliability may be better than the competition.
*I have no idea how the reliability of these two cars differs.
For example, an expensive car's interior panels may be manufactured to a template size ± 0.5mm. In contrast, a cheap car's panels may only be manufactured to ± 2mm, so they cost less to make.
Where this starts to hurt 'premium' brands in the reliability stakes is that because expensive cars are generally less reliable and more expensive to fix than cheap cars, and because premium brands have a disproportionate number of expensive cars, they will tend to score lower in these surveys than mainstream brands, even though class for class (e.g. Porsche Boxter vs Nissan 370z*), the reliability may be better than the competition.
*I have no idea how the reliability of these two cars differs.