Polygon, the Dancing Bear

Occasional notes on politics, history, technology, architecture,
and the life of a county clerk

2002:
May / June / July
August / September / October
November / December

2003:
January / February / March
April / May / June / July
August / September / October
November / December

2004:
January / February / March
April / May / June / July
August / September / October
November / December

2005:
January / February / March
April / May / June / July
August / September / October
November / December

2006:
January / February / March
April / May / June / July
August / September / October
November / December

2007:
January / February / March
April / May / June / July
August / September / October
November / December

2008:
January / February / March
April / May / June / July

Current entries

Monday, January 14, 2008, 5:30 pm

Presidential Primary: Two Dramatic Updates!

First, I was sued this morning. Nothing personal, though. Local Libertarian Party activist and attorney David Raaflaub sued the Secretary of State, the county Election Commission, and the County Clerk, asking for the presidential primary law to be declared unconstitutional, and for an injunction against holding the primary.

Mr. Raaflaub argued that being required to disclose which party primary he is voting in is a violation the Michigan Constitution's guarantee of ballot secrecy.

(I know those of you in states with party registration will find this hilarious.)

This afternoon, Judge Timothy Connors dismissed the suit.

Second, we have heard from the state Bureau of Elections that, due to an error in ballot programming, Uncommitted and Write-in votes will be counted together in many counties. In order to untangle the mess, they will have to do hand counts.

Here's a pretty good, but probably not perfect, list of the affected counties:

Alger, Alpena, Antrim, Arenac, Baraga, Cheboygan, Chippewa, Gogebic, Grand Traverse, Houghton, Iron, Jackson, Keweenaw, Leelanau, Lenawee, Luce, Mackinac, Manistee, Marquette, Montmorency, Ontonagon, Oscoda, Otsego, Presque Isle, Roscommon, Schoolcraft, Wexford.

Bottom line, if this is a close election, it will be a very long night.

....Posted by Lawrence Kestenbaum —

Wednesday, January 9, 2008, 4:44 pm

Today's message to my staff: polls and New Hampshire.

After yesterday's startling result in New Hampshire, inconsistent with all the polls, the news media are asking what went wrong.

Similar questions were raised in the last Detroit mayor's race, when every poll showed that the challenger was way ahead - but the incumbent won by 14,000 votes.

The problem is that accurate opinion polling is harder and harder to do.

The mathematical theory behind polling is that you can talk to a random sample of the entire population, and come up with a pretty good estimate of what the entire population is thinking.

However, if the people in your sample are not a random sample of the relevant population, the whole concept fails.

In past times, almost everybody had a Bell System home telephone, and almost everybody answered the phone when it rang. Polling was comparatively easy then.

Nowadays, with answering machines and Caller ID, we all have more control over who we spend time talking to. Years of unwelcome interruptions from telemarketers have taught us to screen our calls.

More and more people are choosing to delist themselves from the telephone directory, or not to bother with a "land line" at all. For most purposes, these people are lost to polling.

With two-career families, longer commute times, growing use of restaurants for meals, taking kids to more and more extracurricular activities, etc., it's harder for pollsters to find people at home at all.

And there's another problem. A poll-taker who reaches a voter by phone has to cajole participation in a sometimes lengthy interview. Willingness to do this is on the decline. After all, it's an unsolicited call from a stranger asking personal questions - no surprise that people often say "no".

When a population is very heavily polled (like primary voters in New Hampshire?) the weariness with all the calls really sets in. And the more people who decline to participate, the less accurate the results will be.

On the other hand, if political activists and people with strong opinions are more inclined to say "yes" - that means those folks will be overrepresented, and the overall poll result will be skewed.

The growing disconnect between poll results and reality comes back to haunt us as election officials, since the media and the political campaigns rely heavily on polls for news. It's taken for granted that polls are accurate. When the election turns out differently than polls, we may even be accused of some kind of chicanery.

During this frenetic election year, with media outlets constantly trumpeting their latest surveys, don't forget the old campaign cliché: "The only poll that matters is the one on Election Day."

....Posted by Lawrence Kestenbaum —

Wednesday, January 2, 2008, 10:50 pm

Bad Economic News from the Deeds Office

A "sheriff's deed" is recorded when a mortgage has been foreclosed and the property sold at auction. Hence, the number of sheriff's deeds recorded in the Register of Deeds office is a precise indicator of distress among homeowners in that county.

By that standard, things are looking grim here in Washtenaw County, with the number of sheriff's deeds reaching previously unheard-of levels. There were 1,151 in calendar year 2007, compared to 703 in 2006 and 433 in 2005. The month just ended had 147, which is the highest monthly total in years, perhaps ever.

Here's the data by month and year since 2002:

Month	2002	2003	2004	2005	2006	2007
Jan	15	21	25	33	32	88
Feb	9	17	22	28	54	99
Mar	16	10	31	38	59	103
Apr	13	37	23	27	46	72
May	20	11	29	40	53	75
Jun	25	26	13	40	69	76
Jul	19	21	29	41	30	113
Aug	23	29	22	27	68	78
Sep	20	17	21	42	63	85
Oct	19	25	22	42	59	108
Nov	24	20	21	20	93	107
Dec	28	31	30	55	77	147
Totals	231	265	288	433	703	1,151