2000 is waaaayyy enough lol
http://blog.surveymonkey.com/blog/2011/09/15/how-many-people-do-i-need-to-take-my-survey/
Less than 3% margin of error.
Also here a calculator:
http://www.custominsight.com/articles/random-sample-calculator.asp
3% margin of error and 99% confidence if you ask 2000 people even if the total population is 130m people.







