Tuesday, October 4, 2011

The wisdom of crowds?

ABC News reports a poll showing that 55% of people expect the Republican nominee to win next year and only 37% expect Obama to win (8% don't know).  I looked for earlier questions asking people to predict races.  There were quite a few, so to narrow it down I focused on the 1948 race, when Harry Truman won even though all of the polls and pundits said he would lose.  The people didn't do any better.  In February 1947, when asked "Regardless of how you yourself feel--which party do you think will win the Presidential election next year?"  20% said the Democrats, 63% said the Republicans, and 17% didn't know.  In April 1947, the percent saying Democrats was up to 30% and the Republicans fell to 53%.  The last time it was asked (April 1948), 21% picked the Democrats and 57% picked the Republicans. 

After the election, a Roper/Fortune survey asked people who they had expected to win.  Only 19% said Truman, and 77% said Dewey.  It also asked "Before election day, who did you think most of your friends were going to vote for?"  27% said Truman, 40% Dewey, 3% Thurmond, 2% said most were undecided, 9% said evenly divided, and 19% didn't know.  So if people had made the naive assumption that the winner would be the person most of their friends were for, more of them would have predicted it correctly.

PS:  I also ran across a question from 1945 (just after the Labour Party won the election in Britain) that shows the change in political climate  over the years: 
 "Do you think a Labor Party will ever win a presidential election in this country?"  38% said yes, 37% said no, and 25 didn't know. 

No comments:

Post a Comment