The numbers confirm McCain is on the wrong side of history

I have long heard that the Consumer Confidence Index (A measure of how good people feel about the economy) is a good determinant of which party will win the presidential election. Being statistically inclined, I decided to take a look at the data and it turns out that since 1968 (as far back as I could find CCI numbers) it is mostly true.

I looked at the CC index number for September of each year before a presidential election and found that when the CCI is negative, the incumbent party has never won a single time (This event has only happened 3 times though- Bush1 in 1992, Carter 1980, and Ford 1976).

Does this mean that when the CCI is positive in the Sept. before the election that the incumbent party wins? Only about 71% of the time. Gore (2000) and Humphrey (1968) were the only candidates of the incumbent party that lost the election when the CCI was positive (although Gore did win the popular vote). Bush2 (2004), Clinton (1996), Bush1 (1988), Reagan (1984), and Nixon (1972) all won when the CCI was positive in September before the election.

Notably, these contests were all between a challenging party and a person that served as vice president or president in the previous administration. What happens when we have an election between two non-incumbent persons of previous administration (like now, between McCain and Obama)?

In the past Century, this only happened twice: In 1928 between Hoover and Smith and in 1952 between Eisenhower and Stevenson. I could not find CCI values for these years, but I think there are some relevant comparisons between those elections and the McCain-Obama election.

In 1928, Hoover (Republican) won the election largely because he was associated with the economic boom happening at the time (although the Depression would soon follow after his election). Additionally, Republican indiscretions of the previous administration had largely faded away. No other democrats were anxious to even face Hoover. Smith, his opponent was a victim of the anti-catholic and prohibition sentiment at the time. In hindsight, it appears Hoover benefited from a positive CCI economy (the republicans were incumbent).

In 1952, General Eisenhower (republican) won the election against Adlai Stevenson. At the time, Harry Truman was the incumbent democratic administration, but he decided not to run for re-election because his approval ratings were the lowest in American history at the time (only Bush2 and Nixon have surpassed this prestigious record). Additionally, Americans were growing weary of the Korean War and wanted out. However, 1952 was one of the largest growth periods in American history due to the post war industrial boom. It appears that Stevenson (the incumbent) lost the election in a positive CCI economy.

What do the elections of Eisenhower and Hoover tell us?

1. The incumbent looses even with positive CCI when the incumbent party has low approval ratings and is responsible for an unpopular war, even if the economy is good.
   
2. If the economy is good and there are no other distractions (like wars, low approval ratings, controversies), the incumbent party will remain in the White House.
   

These two insights combined with the observations that no incumbent party has won when the CCI is negative is telling. This past August, the CCI was around negative 44. Unless the CCI increases by 100% in the next month (which is very doubtful) the prospects don't look good for McCain.