Thursday, June 29, 2006

Primary Turnout

There has been much said about the moving of the primary from September, which is a bad time to have a primary, to August, which is presumably worse. The reason for the change was to give candidates more time to gather funds and support for the general election in November. Critics have blasted the change, saying that an August primary will draw significantly smaller turnout than a September one, and make life easier for incumbents.

Unfortunately, the historical record doesn't back this up.

In 1970, Connecticut held its first statewide primary. Both Democrats and Republicans held primaries that year, for Senator and Govenor respectively. Interestingly, it was held in August. In 1978, Gov. Ella Grasso faced a bitter primary challenger from her Lt. Gov., Robert Killian. That primary was held in September.

Other September gubernatorial primaries occurred in 1986 and 1990. In 1986, Julie Belaga won the Republican nomination (only to be crushed by William O’Neill); in 1990 Bruce Morrison defeated William Cibes for the Democratic nomination; and in 1994 both parties held primaries, which ended up in John Rowland and Bill Curry facing one another for the first time.

Here are the turnout figures for various statewide races, gathered from newspaper reports:

1970 GOP Governor (August): 33%
1970 Dem Senate (August): 38%

1978 Dem Governor (September): 32% (approx.)

1986 GOP Governor (September): 22%

1990 Dem Governor (September): 20%

1994 Dem Governor (September): 24%
1994 GOP Governor (September): 25%

These figures reflect the percentage of voters registered in the party holding the primary who actually voted in the election. As you can see, the shift from August (1970) to September (1978) didn't really affect turnout for the gubernatorial primary. From 1978 to 1994, there was a significant drop in turnout.

The data here suggests that voters who are willing to turn out for a primary probably don’t care what month it’s being held in. In the 1970s, the percentage of primary voters was in the mid 30s. In the 1990s, that number dropped into the 20s.

It also suggests, although nowhere near as conclusively, that there may be slightly more interest in a U.S. Senate race than in a governor’s race, especially when big issues are at stake. The 1970 primary, like the 2006 primary, was largely about huge issues of war and peace, and attracted a large turnout. Turnout may also be affected by the belief that the party in question can actually win the November election. Turnout for Belaga was low because no one gave the Republicans much of a chance in 1986. In 1994, it seemed like anyone could win.

Given this, it’s possible to guess at what turnout will be on August 8th. Somewhere around 28% of registered Democrats seems reasonable. The race will generate more interest for Democrats than Curry-Larson, and certainly more interest than Morrison-Cibes, which was a foregone conclusion. It’s also probable that the turnout for the upcoming primary will be increased by at least 5%, maybe more, because of the Senate battle. If it were just Malloy and DeStefano, turnout would be in the low 20s. If that.

The conclusion here is that a lot of factors influence primary votes. But a shift from September to August probably isn’t one of them.

Sources

Noel, Don. “Politics: A New National Pastime?” Hartford Courant 19 September, 1994. p. A11.

Treaster, Joseph. “Weicker, Meskill Win in Primaries.” New York Times 13 August, 1970. p.1.

Merry, George. “How Primary Votes Went in New England.” Christian Science Monitor. 14 September, 1978. p.6

7 comments:

Genghis Conn said...

I think 35% is a little optimistic. In 1970, the only year we really have for comparison, turnout for the Senate race was only 5% higher than for the governor's race (primaries held for different parties on different days). So take that 24% and add 5% to it--you get 29%. I don't see an increase of 10% over 1994. The interest just isn't there.

Genghis Conn said...

I think 35% is a little optimistic. In 1970, the only year we really have for comparison, turnout for the Senate race was only 5% higher than for the governor's race (primaries held for different parties on different days). So take that 24% and add 5% to it--you get 29%. I don't see an increase of 10% over 1994. The interest just isn't there.

GMR said...

If Lieberman loses the primary and runs in the general election as an independent, it is going to be very interesting to see what the national Democrats do: the DSCC and the individual Senators in Washington. Whom will they support?

If turnout is less than 25%, they will of course say that the turnout was low and that the vote doesn't reflect the true wishes of the Connecticut voter, and that Lieberman is favored in a three-way race.

If turnout is higher than 30%, then the national Democrats who are backing Joe Lieberman are going to have to tacitly admit that the Democrat party primary voter is out of touch with the mainstream.

I think that most Lamont supporters are going to show up to vote in this primary. Lieberman's support doesn't seem to be as passioned.

Are there certain areas of the state that are more likely to be for Lieberman or for Lamont? Will the governor's race pull any others into the primary?

Genghis Conn said...

Support for Lamont at the convention seemed to be coming from everywhere. I don't think that you can necessarily find a specific area of the state where Lamont or Lieberman will be stronger than any others.

It will be a fun one to map. And oh, yes. It will be mapped.

Genghis Conn said...

DC, it's alphabetical.

Genghis Conn said...

Oops. Fixed, bluecoat.

CC said...

Do any of the figures from the main post come from an INCUMBENT primary?