Inside the Numbers: the 2005 Supervisor's Primary

Let’s take a look at the results of the 2005 Democratic Primary for Town Supervisor. As I mentioned previously, the 2005 contest between Paul Feiner and William Greenawalt was quite close, unlike 2007, when Feiner routed Greenburgh Democratic Party chair Suzanne Berger by a 2:1 margin. I was unsure that the results listed at the Westchester Co. Board of Election website were complete because only 58 of Greenburgh voting districts were listed, compared to 83 voting districts in the primary two years later. The person I spoke to at BOE believed that a number of districts had been consolidated for the primary to save expenses. I was skeptical because I had assumed that the 2005 turnout would be larger than (what I believed to be) the less publicized race in 2007. Without further information, however, I’ll assume that the BOE is correct. What happened in 2005? This was the closest race in Feiner’s many campaigns for Town Supervisor. Paul received 2983 (50.08%) votes, compared to 2786 (47.45%) for his challenger, William Greenawalt (there were 103 blank ballots). This translated to a margin of 197 votes. In 2005, just over 41% of the vote came from the incorporated villages. In contrast with 2007, Feiner performed slightly worse in the villages where he won 49.34 of the ballots. Feiner did, however, have more votes than Greenawalt who trailed Feiner by 25 votes in the Villages, coming in at just over 48%. [If excluding blank ballots, Feiner had 50.5% and Greenawalt 49.5%]. Here are results broken down by village:

VILLAGE	VOTE DIST.	TOTAL	GREEN.	%	FEINER	%
TARRYTOWN	1-7, 55	365	150	41%	207	57%
IRVINGTON	8, 9, 10, 11	391	223	57%	162	41%
DOBBS FER.	12-16, 58	521	221	42%	279	54%
HASTINGS	18-23, 52	704	318	45%	368	52%
ARDLSEY	24, 25	327	222	68%	102	31%
ELMSFORD	26, 28	110	34	31%	75	68%

Feiner did slightly better in unincorporated Greenburgh ("UG" - I like that acronym) where he received 1790 votes (51.8%) versus 1618 for Greenawalt (46.8%). If taking out the blanks, Feiner goes up to 52.5% and Greenawalt down to 47.5% in UG. The question I struggle with is how to analyze the UG vote. The “town outside” includes residents spread across numerous distinct neighborhoods in eight different school districts with addresses that can be more than 8 miles apart. The U.S. Census Bureau recognizes three “census designated places” in UG: Fairview, Hartsdale and Greenville (Edgemont). These CDPs, however, are of limited usefulness. The borders designated by the Census Bureau for these areas are quite restricted and don’t match up with the way these places are understood by their residents. 
See: http://www.maptechnica.com/us-city-boundary-map/city/Hartsdale/state/NY/cityid/3632523
For example, the Hartsdale CDP does not include the Poet’s Corner neighborhood, although, I believe, people who live there consider themselves to live in Hartsdale. Certainly the location of the Hartsdale sign on W. Hartsdale Ave. suggests that's the case. For the most part, people in our area of Westchester Co. tend to identify their community by their school district. This obviously applies to Edgemont which has constructed a strong identity centered on their school system (and high property values and insane taxes that come with it). But most of UG lies in the Greenburgh Central school district ("GC") which is definitely not the focus of communal identity. You don't see Woodlands stickers decorating cars the way you see Raiders and Panthers decals. But trying to apply consistent factors to define neighborhoods devised of unrelated parcels leads to a deadend. Consequently, for purposes of this voting analysis, I’m creating an expanded definition of the CDPs Hartsdale and Fairview. For Hartsdale, I’m using the large portion of the 10530 zipcode that lies inside the GC school district. For Fairview, I’m using the northern half of the GC school district. I’m separately listing the portions of uninc. that lie outside of GC and identifying them by their school district.

SCHOOL DIST.:	VOTING DIST.	TOTAL	GREEN.	%	FEINER %
EDGEMONT	33, 34, 35, 49	650	422	65%	223	34%
GC (HARTS):	36-41, 56, 57, 71	962	462	48%	488	51%
GC (FAIR):	42, 44, 45, 46, 64	759	304	40%	438	58%
POCANTICO	29, 68	154	62	40%	89	58%
VALHALLA	47, 48	280	98	35%	179	64%
ARDSLEY:	32, 43, 50	429	187	44%	238	55%
ELMSFORD	51	105	45	43%	58	55%
TARRY/ELM:	30	63	17	27%	46	73%
IRV/ELM:	31	52	21	40%	31	60%

NOTES: Parts of Dist. 50 appear to also lie inside GC. According to the map I received from the BOE, Dist. 30 is split between the Tarrytown SD and Elmsford SD while Dist. 31 is similarly split between split between Irvington SD and Elmsford. I have no idea if the residents in those districts predominantly identify with a particular town/school district. Also, I’m assuming that the voting districts in the 2005 primary results more or less correspond to the 81 voting districts that appear on the BOE map. There may be some discrepancies.

 A few things jump out: notice the contrast between Irvington and its neighboring villages to the immediate north and south. Look at the mirror image percentages between Ardsley and Elmsford. As far as UG: Edgemont is the only district that Feiner lost (and he lost big). Contrast Edgemont, however, with Feiner’s popularity in the Valhalla school district neighborhoods. I don't understand at all the gap between Ardsley village's vote and the UG neighborhoods that are in the Ardsley SD. Next, we’ll compare these numbers with 2007 and examine how Feiner responded to these results and brilliantly adapted: just two years later, he triumphed easily.