First, that's a survey (no actual usage data). Secondly, that differs from what consumers with a choice use. Many poor don't have credit cards; many more have impaired credit.