Effect of Presidential Coattails on the Senate

In 2016, both the White House and control of the Senate are up for grabs. A key question is: "How much do the two contests track each other?" More specifically, when a party wins the White House, does the presidential candidate also have coattails to help his party gain seats in the Senate? To examine this question, we have collected the data back to the Civil War (specifically, 1864) to see how well the presidential election correlates with gains or losses in the Senate. Of the 34 presidential elections where one of the two major parties gained in the upper house, the President pulled in some seats for his party 20 times and saw his party lose seats 14 times. If we examine the 20 times the President's party also won in the Senate, the mean gain was 5.3 seats. On the other hand, when the President had negative coattails, the effect was weaker, with his party losing 2.7 seats on average. If we average all 38 elections, including the four where neither major party gained seats, the President pulls in 2.0 seats on average. So, if Hillary Clinton wins the presidency, statistically speaking, she is likely to help the Democrats get a net of two seats in the Senate—just shy of the four seats she would need for the Democrats to take control. This would lead to at least two more years of divided government, the same as now. Of course, she is unlikely to hit the exact average if she wins, and presidential victors have seen their party gain as many as 13 seats (1932) and lose as many as 7 (1876), or as many as 4 (2000) if we limit ourselves only to those contests held since the direct election of senators began in 1916. The takeaway is that if Clinton wins, the Democrats have a shot at controlling the Senate, but that is far from a certainty, so the Senate will be a big battleground next year.

Year Presidential winner Senate change Coattails?
1864 Republican R+4 Yes
1868 Republican R+2 Yes
1872 Republican D+2 No
1876 Republican D+7 No
1880 Republican R+6 Yes
1884 Democrat R+4 No
1888 Republican R+1 Yes
1892 Democrat D+5 Yes
1896 Republican No change  
1900 Republican D+5 No
1904 Republican R+5 Yes
1908 Republican D+3 No
1912 Democrat D+2 Yes
1916 Democrat No change  
1920 Republican R+9 Yes
1924 Republican R+7 Yes
1928 Republican R+5 Yes
1932 Democrat D+13 Yes
1936 Democrat D+2 Yes
1940 Democrat R+1 No
1944 Democrat No change  
1948 Democrat D+9 Yes
1952 Republican D+2 No
1956 Republican D+2 No
1960 Democrat No change  
1964 Democrat D+3 Yes
1968 Republican R+5 Yes
1972 Republican D+2 No
1976 Democrat D+1 Yes
1980 Republican R+11 Yes
1984 Republican D+2 No
1988 Republican D+1 No
1992 Democrat R+1 No
1996 Democrat R+2 No
2000 Republican D+4 No
2004 Republican R+4 Yes
2008 Democrat D+11 Yes
2012 Democrat D+2 Yes