Question 1193063
population parameter should be % of people of each political persuasion in the city/mayoral area. But 
parameter of interest is presumably voters who were democratic, if I can assume "leaning democratic"=democratic.
Sample is those who responded. Note: who responded. 
Statistic is the 38.2% of people, the measure of the sample.
Margin of error is how much uncertainty. To deal with that, one would need to know the number of respondents, and that term alone makes "random sample" suspect. 
Because we do not know margin of error, we do not know a 95% confidence interval. Confidence is how certain we are of where the parameter lies.
-
If one were to say that 38.2% of the 814 voters in the sample leaned democratic, then we could have a margin or error and a 95% confidence interval.

Caution always when interpreting data where it says "respondents."  There are plenty of instances where respondents are not at all representative of the population.