You compare the average bicycle fatalities in states with and without the helmet law in place. You observe states with more fatalities are less likely to have the helmet law. Is this convincing evidence of a causal link?