# Political Science Methods Economists Forecast Model

Read the methodology descriptions for the FiveThirtyEight and Economist election forecasts:

https://fivethirtyeight.com/features/how-fivethirtyeights-2020-presidential-forecast-works-and-whats-different-because-of-covid-19/

https://projects.economist.com/us-2020-forecast/president/how-this-works

Some of the details are technical, but they also provide a lot of description of the inputs (variables and polls) that they take into account in their models and the assumptions that they make in generating their forecasts. After reading the two descriptions, which model do you believe is likely to produce a more accurate forecast? Point to specific differences between the two models.

2. What is your prediction of the final electoral college result? How many electoral college votes will Donald Trump and Joe Biden each receive?

3. Here is a survey result from late September for the Kansas Senate race. It reports the “general design effect” and the full margin of error? Using the method we learned in class to calculate a sampling error for a mean, what is the portion of the margin of error that comes from sampling error for this survey (don’t just subtract the general design effect from the margin of error)?

https://civiqs.com/documents/Civiqs_DailyKos_KS_banner_book_2020_09_dw29u7.pdf

4. Which candidate do you predict will win the Kansas Senate Election? What share of the vote will Roger Marshall receive and what share of the vote will Barbara Bollier receive?

5. What is your prediction for the distribution of seats in the U.S. Senate between Republicans and Democrats? Count any independents in the party that they caucus with (i.e. Bernie Sanders is an Independent, but he caucuses with the Democrats in the Senate).