by Theodore de Macedo Soares

According to the exit polls conducted by Edison Research, the Democratic Party Senate candidates won three Senate seats that went to the Republican candidate in the computerized vote counts. With these additional seats, the Democratic Party would have regained control of the Senate—51 seats to 48 for the Republican Party and one Independent (caucuses with the Democratic Party).

Edison Research conducted exit polls in only 28 states in the 2016 general elections.  Senate seats were contested in 21 of these states.  The contests in 20 states featured races between Democratic and Republican Party candidates. Discrepancies between the official vote count and the exit polls favored the Republican Party candidate in 17 states. In 12 of these states the discrepancies were beyond the exit poll margin of error. See table and notes below.

The link Exit-Polls_2016-US-Senate-Elections.zip  contains the exit polls downloaded from CNN shortly after the polls closed in each state.  If using these files please credit www.tdmsresearch.com  and Theodore de Macedo Soares as the source and most importantly cite the article that shows the results of these exit polls: https://tdmsresearch.com/2016/11/15/2016-us-senate-elections/

[1] Exit polls (EP) conducted by Edison Research and published by CNN shortly after the closing of polls for the state and downloaded by TdMS. Jonathan Simon downloaded MO EP. Edison Research conducted exit polls in only 28 states including 21 states with senate races. Exit poll results are derived from the gender category–the proportion of men and women voting for each candidate.

As these first published exit polls are subsequently altered/adjusted to conform with the unverified computer vote counts, the discrepancies shown above are adjusted to near zero in the later and final published EPs. As noted above, such alterations apply prior to these first published polls in the states marked by an asterisk. These states embody two time zones with the bulk of the state in the earlier zone and poll closing time. Indiana, in this election, is a notable example: With the bulk of Indiana’s vote, coming from the earlier time zone, at 9.7% in favor of the Republican, it is possible that the original exit polls showed a Democratic candidate win by as much as 3.8% (5.9% published exit poll result less 9.7% incoming computer count result).

[2] Official FEC vote counts published December 2017. Available at: https://transition.fec.gov/pubrec/fe2016/federalelections2016.pdf

[3] The margin columns subtracts the Democratic Party Senate candidate totals from the Republican Party Senate candidate.  A Republican Party candidate win is shown by a positive sign and a Democratic Party candidate win by a negative sign. Due to rounding some totals may appear slightly inconsistent.

[4] Note that the Margin of Error (MOE) is for the differences between the two candidates (at 95% CI). This MOE is about double the usual MOE for each candidate. MoE calculated with multinomial formula discussed in sections 2 and 4 in:  Franklin, C. The ‘Margin of Error’ for Differences in Polls. University of Wisconsin, Madison, Wisconsin. October 2002, revised February 2007. Available at:  https://abcnews.go.com/images/PollingUnit/MOEFranklin.pdf

[5] In California, the two candidates were both Democrats: Kamala Harris (represented in the Democratic Party columns) and Loretta Sanchez. http://www.weeklystandard.com/californias-no-republican-senate-race/article/2004578

Update: November 21, 2016. Vote counts updated. Only California vote count changed slightly.

Update: August 14, 2019. As the 2020 elections approaches table was updated with additional information on the states with duo poll closing times.