[1] is better interpreted as 40% of people who are willing to work for less than minimum wage answering marketing surveys don't have $400 for an emergency. It could still be true, but the bias in the methodology is obvious.