Senator Alberta Darling
Recall Petition Signature Analysis
Analysis of recall signatures collected from Mar. 3, 2011 to Apr. 21, 2011.
Purpose
With the onslaught of recall elections in 2011 and recent concerns regarding duplicate and invalid signatures, the purpose of this project is to examine an actual recall petition drive to uncover any trends, good or bad, that can be learned from it.
This project was carried out entirely by John Von Haden, a private citizen (and constituent of Senator Darling) residing in Menomonee Falls. However, the entire project was conducted without Senator Darling’s knowledge/involvement and was NOT done in conjunction with any tea party group, political group, campaign, nor any elected official.
The data analyzed comes entirely from public information posted on the Government Accountability Board (GAB) website) at http://gab.wi.gov/node/1655 . On their website, the GAB has made the recall petition forms available to the public via scanned images in PDF files.
The results of this project do NOT refute the GAB finding that enough signatures were collected to force a recall election to take place.
Concerns
Missing Information
The results in this report are skewed by some missing information in the PDF files.
The GAB website has links to 92 files, each containing scans of 50 petition forms per file. The GAB gave each page a unique number, starting with 1A, 1B, 1C, 1D, 2A, 2B, 2C, 2D, 3A, 3B, and then 4 through 4700.
The following pages are missing from the PDF’s: 45, 116, 121, 122, 165 thru 168, 212, 488, 493, 863, 1084, 1713, 1722, 1832, 1833, 2799, 4451 thru 4550, and 4603 thru 4624. The group 4451 thru 4550 are two entire PDF files for which there is no link on the website. In all, 140 of 4700 pages are missing.
Additionally, each page is understandably cropped to remove personal information on the right side (email address and phone number). However, either through a poor scanning procedure or an overzealous GAB staff member, eighty eight pages were unnecessarily cropped further to remove the following information:
- The printed name of the signer
- ALL information pertaining to the circulator
- ALL information for the 9th and 10th signatures on each page (generally blank anyway, but not always)
The cropped pages are pages 2826 thru 2850, 2924 thru 2950, 3597 thru 3600, and 3869 thru 3900.
Duplicate Page
This only happened one time, but is odd. The petition with GAB page number 2324 is a photocopy of page 2340. Thus one petition page was submitted twice, however, the GAB staff did not detect this, giving the pages unique numbers and allowing them both to be counted.
Results
Duplicate signatures
Below are counts of duplicate signatures found (not all were found)
| Constituents who signed twice | 655 |
| Constituents who signed a third time | 49 |
| Constituents who signed a fourth time | 4 |
| Non-constituents who signed twice | 17 |
| Total | 725 |
Note: “Constituents” above (for lack of a better word) does not indicate that the persons live at the addresses they gave nor that they are eligible voters, only that the address they gave is within Senator Darling’s district (as it existed in the spring of 2011).
On page 2744, someone was so bold as to sign twice on the SAME petition form.
Plotting the duplicate signers’ addresses on a map shows that duplicates occur predominantly in Shorewood, Whitefish Bay, Glendale and Brown Deer. Three particular neighborhoods stand out as being heavily canvassed multiple times causing approximately 50 to 100 duplicates in each neighborhood.
1. North of Brown Deer Road and east of 76th Street – among others includes sixty one duplicates on Marien St. to 75th Street to West Glenbrook Road.
2. North of Brown Deer Road and just west of Northridge Mall – among others includes eleven duplicates on 85th Street, fifteen duplicates on Joyce Ave, and fourteen duplicates on Fairy Chasm Drive.
3. North of Brown Deer and from 91st street to the east, among others includes twenty one duplicates on Swan Road, thirty four duplicates on 95th Street, nineteen duplicates on W. Allyn St., and twelve duplicates on Jolena Lane to 94th Street.
These numbers are rather high (about 2.4 percent) given that there was little if any public discussion regarding duplicates in the spring of 2011, whereas news accounts in December 2001 suggest that duplicates were being encouraged (with no repercussions) during the Governor recall petition drive.
Ficticious names
“Adolf Hitler” signed on page 3504, line 6, listing an address of ” 1313? W. Glenview? “ in Berlin. Although the GAB has pointed out that this one was rejected because Berlin is not in the Senator’s district, for the statewide Walker/Kleefisch recalls, one could argue that Berlin, Wisconsin is within the state of Wisconsin.
“Dick Head” signed on page 1254, line 9, listing an address of ”5 Dick Head Lane” in Whitefish Bay.
“Union Thugs Suck” signed on page 2863, line 10, listing an address of “No Frigging Way”.
Non-constituents
Approximatly 2,600 signatures gave addresses which are not within Senator Darling’s district. One was out of state (Chanhassan, Minnesota).
Miscellaneous - Electors
115 signers did not put a date.
86 signers gave only a municipality, without giving their street address.
174 signers gave a street address, but did not list their city/municipality
59 signers listed a date that was PRIOR to one of the prior signers on the same page.
11 signers gave their first name or last name only, but not their full name.
36 signers gave their printed name only without a signature.
9 signatures listed dates prior to Mar. 3, 2011
Miscellanous – Circulators
27 signatures were on forms on which the circulator did not date the form
9 signatures were on forms on which the circulator did not even list their name
5 signatures were on forms on which the circulator did not give their street address
201 signatures were on forms on which the circulator did not list their city/municipality
119 signatures were on forms on which the date given by the circulator is PRIOR to the date of the signer.
Signature counts (by day)
| Week of… | Sun | Mon | Tue | Wed | Thu | Fri | Sat |
| Mar. 3 – 5 | 48 | 16 | 3396 | ||||
| Mar. 6 – 12 | 3232 | 500 | 625 | 316 | 757 | 1156 | 2284 |
| Mar. 13 – 19 | 2944 | 765 | 788 | 798 | 420 | 469 | 2252 |
| Mar. 20 – 26 | 329 | 328 | 112 | 74 | 134 | 138 | 1287 |
| Mar. 27 – Apr. 2 | 696 | 106 | 223 | 152 | 198 | 38 | 778 |
| Apr. 3 – Apr. 9 | 265 | 116 | * 4816 | 55 | 199 | 183 | 486 |
| Apr. 10 – Apr. 16 | 270 | 202 | 171 | 439 | 299 | 231 | 225 |
| Apr. 17 – Apr. 21 | 460 | 264 | 20 | 227 | 8 |
Note that by the fourth week the weekday average was only in the 100 to 200 range and the Saturday/Sunday average was only in the 200 to 500 range. It is clear that not enough signatures would have been collected were it not for the timing of the spring election (*) on Tuesday, April 5, when petitioners were able to collect about 4,800 signatures by standing outside the polls. Incidentally, about 150 duplicates and about 180 non-constituents signed on that day (April 5).
Signature counts (by circulators’ residences)
| Circulator Residence | Within Senate District | Outside of Senate District |
| Dane county | 2060 | |
| Dodge county | 258 | |
| Jefferson county | 55 | |
| Kenosha county | 159 | |
| Manitowoc county | 29 | |
| Marathon county | 39 | |
| Milwaukee county | 11831 | 9526 |
| Ozaukee county | 1147 | 173 |
| Racine county | 311 | |
| Rock county | 145 | |
| Walworth county | 27 | |
| Washington county | 988 | 81 |
| Waukesha county | 1358 | 1253 |
| Other counties | 24 | |
| Illinois | 157 | |
| Other states (CA, FL, MI, NY, OR, PA, WA) | 49 | |
| Undetermined (738) | ||
| Totals | 15324 | 14427 |
The undetermined residences are caused by (1) circulators who did not list their address and/or municipality and (2) the 88 pages on which the circulator information was cropped out of the digital images.
This table includes all signatures (duplicates, rejected and valid signatures)
During the first weekend (Mar. 5 and 6), the percentage of signatures collected by constituents was about 59 percent and even higher during subsequent weekdays (due to the inconvenience of travel for non-constituents on weekdays).
This trend lasted for only two weeks. During the third week (starting on Saturday, March 19), the split was generally about 50/50.
During the last few weeks (March 26 through Apr. 20), with a few exceptions, the out of district circulators collected more signatures than constituents (one exception was a 50/50 split on Apr. 5, the day of the spring election).
Overall, out of district circulators collected about half (48 percent) of the signatures.
Conclusion
The five most successful signature collection dates were the first two Saturdays and Sundays and the Apr. 5 spring election. The enthusiasm of constituents seems to have dropped off after the second Sunday.
Major success factors:
1. Out of district circulators were crucial to this recall, especially after the third week.
2. The collection of 4,800 signatures (most outside polling places) during the April 5 spring election far exceeded all other days.
Without either of the above factors, not enough signatures would have been collected to force a recall election.