This might very well be enough. For 35 million people 2% error is 2400 people. So I think we can say it is 46-50% of the population does not have $200. Howev

This might very well be enough. For 35 million people 2% error is 2400 people. So I think we can say it is 46-50% of the population does not have $200. However if the samples were already biased into certain groups or not even Canadian, then it gets thrown off.