Screenshot from crimemapping.com (Image links to crimemapping.com)

CrimeView Dashboard is an Omega Group software product used internally by the Ann Arbor police department to help solve crimes. CrimeView is far more robust than Omega Group’s publicly available, web-based crimemapping.com, which provides citizens with crime maps based on data from the most recent six months.

In a phone interview with Chris Baldwin, Omega Group account manager for Ann Arbor, he told The Chronicle Omega views the company’s relationship with AAPD as a partnership. The publicly available crimemapping.com tool, Baldwin said, is relatively lightweight compared to the tools available in CrimeView, but is still useful for educating and informing the public.

As one example of the difference, crimemapping.com gives locations by street blocks, whereas the AAPD has access to the exact street address.

Through CrimeView, AAPD officers also have access to data on all calls for service, not just on the verified incidents. In a telephone interview, AAPD deputy chief Greg Bazick explained that a call for service is not automatically logged as an incident. The call initiates the report, but then it’s reviewed and validated – which means that, in the view of AAPD staff, an incident meeting the description took place. It’s those validated incidents that appear on crimemapping.com

CrimeView allows officers to set up and save standard data queries so that trends and patterns can be spotted, which can drive decisions about how to deploy resources with an eye toward solving crimes. Bazick described how CrimeView played a supporting role in a recent series of home invasions (breaking and entering). That effort involved sharing information with Pittsfield Township,, which also uses the CrimeView Dashboard.

CrimeView was used to plot reported crimes by date, time and location. Because AAPD had the suspect’s home address, known associates, and knew the suspect had no regular access to a vehicle, AAPD could identify the pattern of home invasions geographically, by day of week, and hour of the day, Bazik said. While CrimeView didn’t itself solve the crime, it saved a lot of time compared to the effort it would have taken to compile the information manually.

Bazik described the AAPD as still in the early stages of learning and using the CrimeView Dashboard tool. It’s part of the AAPD’s broader effort to take a data-driven approach to decision making.

That broader effort includes a project that takes a digital approach to officer activity reports. Having easily analyzable data about how officers are spending their time will allow the AAPD to provide a metric that’s important for measuring success in public safety as defined by the Ann Arbor city council. The description of public safety success, developed by the council at a December 2012 retreat, reads in part: “Police officers have between 25-30% of their time available for proactive policing.”

The AAPD started using electronic data sheets for officer activity logs around the beginning of the year. At the council’s July 15, 2013 meeting, AAPD chief John Seto told councilmembers that the department was doing the initial analysis of data that’s been recorded so far. Responding to an emailed query in late August, Seto wrote to The Chronicle that the analysis of timesheets would be reported to the city council sometime in the near future.

At the July 15 council meeting, Seto also briefed councilmembers on the somewhat downward trend for overall crime for the first six months of 2013, compared with the first six months of 2012. For example, based on the publicly available data from crimemapping.com for the additional month of July, The Chronicle counted 249 breaking-and-entering (forced and unforced entry) incidents from January-July 2013, compared to 295 for the same period in 2012.

Slightly more detail on the breaking-and-entering data is provided after the jump.

Ann Arbor: Breaking and Entering

The earliest Ann Arbor data available through crimemapping.com was from February 2011. One limitation on crimemapping.com is that only data from the most recent six months is available. Chris Baldwin, Omega Group account manager for Ann Arbor, told The Chronicle that the six-month window is based on a desire to have a standardized window for all agencies participating in crimemapping.com. In Michigan, that’s around 30 different departments.

However, The Chronicle has routinely downloaded the data available on crimemapping.com and added some data fields to facilitate analysis in common spreadsheet pivot tables (e.g., a “COUNT” flag). [.csv file of Ann Arbor crimemapping.com data 02.01.2011-07.31.2013]

The series of “home invasions” described by the AAPD deputy chief fall into the general category of “breaking and entering” crimes. Those are coded as at least five different types in the crimemapping.com dataset:

• B&E – Burglary – Forced Entry – Non-Residence
• B&E – Burglary – Forced Entry – Residence – Home Invasion
• B&E – Burglary – No Forced Entry – Non-Residence
• B&E – Burglary – No Forced Entry – Residence- Home Invasion
• Burglary – Unoccupied Building or Other Structure

To graph and map the data, The Chronicle filtered it based just on forced versus unforced entry (instead of residence versus non-residence). Geocoding was done with GPS Visualizer and the mapping was done with geocommons.com.

B&E bar graph – February 2011 through July 2013.

B&E Map Geocoding of the address blocks – which is the more general form available through crimemapping.com – is not as reliable as with exact addresses. Some of the crime locations shown may not be accurate.

The Chronicle could not survive without regular voluntary subscriptions to support our coverage of public bodies like the city of Ann Arbor. Click this link for details: Subscribe to The Chronicle. And if you’re already supporting us, please encourage your friends, neighbors and colleagues to help support The Chronicle, too!

Primary elections in the city of Ann Arbor this year fall on Tues. Aug. 2.

Ann Arbor residents are represented by the mayor and 10 other elected members on the city council – two for each of the city’s five wards. Each year, one of the pair of councilmembers stands for re-election to a two-year term. This year, three members of the currently all-Democratic council have contested races in the primary, which falls on Tuesday, Aug. 2.

In Ward 5, incumbent Mike Anglin is challenged by Neal Elyakin. In Ward 3, incumbent Stephen Kunselman is challenged by Marwan Issa and Ingrid Ault. And in Ward 2, incumbent Stephen Rapundalo is challenged by Tim Hull.

Each of the city’s five wards is divided into precincts.

In this article, The Chronicle takes a look at the incumbents’ performance in past elections, mapped out by precinct. Some descriptive generalizations are readily apparent in the data – the strength of incumbents has not been uniform across their respective wards.

And in some cases, it’s possible to offer a speculative analysis that could account for some of those patterns.

Ward 5

Ann Arbor’s Ward 5 covers the central, western part of the city.

Ward 5: August 2009 – Anglin, Rosencrans

Map A shows results for the 2009 Democratic primary election for Ward 5 between incumbent Mike Anglin and challenger Scott Rosencrans. [Washtenaw County clerk's office results] Anglin successfully defended his seat, winning the election with 65% of the vote (1,301 votes to 694). This map plots Anglin’s relative success in the 11 precincts of Ward 5. The pushpins indicate where each candidate lives, and the black lines mark the individual precinct boundaries of the ward.

Map A. Anglin-Rosencrans 2009 (Image links to dynamic Google Map)

The shade of blue corresponds to Anglin’s relative success in a precinct. A darker shade of blue means Anglin received a higher percentage of votes in that precinct. For example, Anglin performed strongest in precinct 5-10, located in the northwestern corner of the ward, where he received 78.9% of votes. So 5-10 is the darkest shade of blue on the map. For those lighter blue parts of the map, Anglin received a lower percentage of votes. He was least strong in precinct 5-8, receiving 52.5% of votes. So precinct 5-8, located directly below 5-10 in the western part of the ward, is the lightest shade of all the precincts.

Ward 5: August 2007 – Anglin, Woods

In the 2009 election, Mike Anglin was defending his seat on the council. But two years earlier in 2007, he was the challenger to Wendy Woods, who had been serving Ward 5 since 2001. Map B shows results for the 2007 Democratic primary election for Ward 5 between Anglin and Woods. Anglin defeated the incumbent, winning the election with 60% of the vote (989 votes to 671). [Washtenaw County clerk's office results] Like Map A, this map plots Anglin’s relative success in the 11 precincts of Ward 5. The pushpins indicate where each candidate lives, and the black lines mark the individual precinct boundaries of the ward.

Map B. Anglin-Woods 2007 (Image links to dynamic Google Map)

In the 2007 Democratic primary, Anglin performed strongest in precinct 5-2, the precinct in which he lives, receiving 66.9% of votes. So 5-2 is the darkest shade of blue on the map. He was weakest in precinct 5-7, receiving 21% of votes. Precinct 5-7, farthest south, is the lightest of all the precincts.

Ward 5: Descriptive Generalizations

In comparing the two elections, Anglin was not consistently strongest or weakest in the same precincts. For example, in the 2009 election against Rosencrans, Anglin won 78.9% of the vote in precinct 5-10, making it his strongest precinct. But two years earlier, against Wendy Woods, 5-10 was his second-weakest precinct, after 5-7. In fact Anglin lost precinct 5-10 against Woods, garnering only 44.6% of the vote.

Even though 5-7 was Anglin’s weakest precinct against Woods (21%), it was one of his three strongest precincts against Rosencrans. The relatively low total number of votes in that precinct (29 in 2007 and 50 in 2009) means definitive conclusions about reasons for the flip are difficult. But City-Data.com shows one-third of residents of that area of the city to be African American, compared with 6.5% citywide. To the extent that African American neighborhoods might tend to support an African American candidate, like Woods, her strong showing there could partly be explained.

In the 2009 contest against Rosencrans, the flip of 5-7 to one of Anglin’s strongest precincts could be accounted for by the overlap of an active neighborhood association with the precinct – Friends of Dicken Woods. The leadership of the association includes Jack Eaton, who supported Anglin’s campaign.

Where each candidate lived had a discernable impact in some cases, but did not seem to be a key factor determining the outcome of the election. In 2007, Anglin did fare best in his home precinct, receiving 66.9% of votes. But Anglin’s performance was about as strong in many other precincts. For example, he received 66% of the vote in precinct 5-11.

Anglin’s opponents also did not appear to enjoy enough of an advantage in their home precincts to have a clear effect on election results. In 2009, Anglin was weakest in precinct 5-8, where his opponent Scott Rosencrans lived – but Anglin still carried that precinct with 52.5% of the vote. And in 2007, one of Anglin’s strongest precincts was the home precinct of his opponent, Wendy Woods – in 5-5 Anglin received 63% of the vote.

Ward 3

Ann Arbor’s Ward 3 covers a wedge of the city starting near the center and widening out towards the southeast.

Ward 3: August 2009 – Kunselman, Bullington, Greden

Map C shows results for the 2009 Democratic primary election for Ward 3, contested by Stephen Kunselman, LuAnne Bullington, and incumbent Leigh Greden. At that time Kunselman had served on city council from 2006-2008, before Christopher Taylor defeated him in the 2008 Democratic primary.

Map C. Kunselman-Bullington-Greden 2009 (Image links to dynamic Google Map)

In 2009, Kunselman narrowly won back his seat, winning 511 votes (36.45%) to Greden’s 505 and Bullington’s 382. [Washtenaw County clerk's office results] The map plots Kunselman’s relative success in the nine precincts of Ward 3.  The pushpins indicate where each candidate lives, and the black lines mark the individual precinct boundaries of the ward.

The shade of purple corresponds to Kunselman’s relative success in a precinct. A darker shade of purple means Kunselman received a higher percentage of votes in that precinct. For example, Kunselman performed best in precincts 3-1 and 3-2, the two side-by-side precincts in the northwestern tip of the ward. Kunselman received 66% of the vote in the two precincts combined. [Note: Because the two precincts were recorded and reported together, the same percentage is assigned to both precincts.]

So precincts 3-1 and 3-2 are the darkest shades of purple on the map. For those lighter purple parts of the map, Kunselman received a lower percentage of votes. He performed weakest in precinct 3-3, receiving 27% of votes. So precinct 3-3, the precinct where LuAnne Bullington lives, is the lightest of all the precincts.

Ward 3:  August 2008 – Kunselman, Taylor

In the August 2008 Democratic primary, Kunselman was challenged and defeated by Christopher Taylor, receiving only 805 votes (36.7%) to 1,387 for Taylor. [Washtenaw County clerk's office results] Map D shows results for the 2008 Democratic primary between Kunselman and Taylor.

Map D. Kunselman-Taylor 2008 (Image links to dynamic Google Map)

Like Map C, the shade of purple corresponds to Kunselman’s relative success in a precinct. A darker shade of purple means Kunselman received a higher percentage of votes in that precinct. So the darkest shade of purple on the map is in Kunselman’s home precinct, 3-6, where he received 57% of the vote. Other precincts in the southeastern end of the ward are also relatively darkly shaded. Kunselman received 55% of the vote in 3-5 and 52% in 3-9.

The map shows the lightest shades of purple in precincts 3-3 and 3-4, where Kunselman polled only 19% and 25%, respectively.

Ward 3: August 2006 – Kunselman, Ralph, Meyers

Map E shows results for the 2006 Democratic primary election in Ward 3 between Stephen Kunselman, Alice Ralph, and Jeff Meyers. Kunselman narrowly won the three-way race, getting 663 votes (36.8%) to Ralph’s 634 and Meyers’ 505. [Washtenaw County clerk's office results] The three were contesting the seat left open by Jean Carlberg, who decided not to seek another term.

Map E. Kunselman-Ralph-Meyers 2006 (Image links to dynamic Google Map)

Like Maps C and D, the shade of purple corresponds to Kunselman’s relative success in a precinct. A darker shade of purple means Kunselman received a higher percentage of votes in that precinct. Kunselman was strongest in precinct 3-6, his home precinct, receiving 54.9% of the vote. So precinct 3-6 is the darkest shade of purple on the map. He performed the weakest in precinct 3-3, receiving 20.7% of votes. So precinct 3-3, the precinct where both Alice Ralph and Jeff Meyers live, is the lightest of all the precincts.

Ward 3: Descriptive Generalizations

In comparing the three elections, Kunselman’s weakest precinct was consistently 3-3. In every election, precinct 3-3 was home to at least one of Kunselman’s opponents – in 2006, both opponents lived in 3-3. That may help to explain Kunselman’s weak performance in that precinct.

Kunselman’s very strongest precinct has varied. In the combined results of 3-1 and 3-2 in 2009, Kunselman polled 66%. Even though those two precincts combined offered only 68 total votes, the election was decided by just 6 votes. So the strength of his showing there, in predominantly student districts, may have been decisive in his victory.

Consistently strong for Kunselman has been his home precinct of 3-6. He’s carried the 3-6 precinct in all three elections, even in 2008, when he lost 3-3 with only 19% of the vote (in a two-person race) in that precinct.

The contrast between 3-6 and 3-3 is likely not just a matter of their geographic proximity to Kunselman and his opponents’  home neighborhoods. According to City-Data.com, the median household income in 2009 for the North Burns Park chunk of precinct 3-3 was $109,000, and for another chunk of it, south of Burns Park, the median household income that year was $91,000. In contrast, precinct 3-6 includes large blocks where the 2009 median household income ranges between $40,000-$55,000.

In this year’s candidate forums, Kunselman has spoken as a point of emphasis about his support throughout the entire ward.

Ward 2

Ann Arbor’s Ward 2 covers the northeast part of the city.

Ward 2: November 2007 – Rapundalo, Amonsen

The most recent Ward 2 election in which incumbent Stephen Rapundalo was challenged took place in the general election of 2007 – he wasn’t challenged in the Democratic primary that year, nor was he challenged in 2009. Ed Amonsen was a write-in candidate for the 2007 general election.

Map F. Rapundalo-Amonsen 2007 (Image links to dynamic Google Map)

Map F shows results for the 2007 general election for Ward 2 between Stephen Rapundalo and Ed Amonsen. Rapundalo successfully defended his seat, winning the election with 51.6% percent of the vote – 843 votes to 790. [Washtenaw County clerk's office results] This map plots Rapundalo’s relative success in the nine precincts of Ward 2. The pushpins indicate where each candidate lives, and the black lines mark the individual precinct boundaries of the ward.

The shade of green corresponds to Rapundalo’s relative success in a precinct. A darker shade of green means Rapundalo received a higher percentage of votes in that precinct. For example, Rapundalo performed best in precinct 2-2, the small precinct in the western part of the ward. Rapundalo received 100% of votes – but only two people voted in that precinct.

A more relevant example of his strong performance is precinct 2-9, located in the northern tip of the ward, where he received 80.3% of votes. Of the precincts that had any kind of voter turnout, 2-9 is the darkest shade of green on the map, followed by 2-6. For those lighter green parts of the map, Rapundalo received a lower percentage of votes. He performed weakest in precinct 2-5, receiving 25.1% of votes. So precinct 2-5, located in the southeast corner of the ward, is the lightest shade of all the precincts.

Ward 2: November 2005 – Rapundalo, Bourque

While Stephen Rapundalo successfully defended his seat in 2007, he first ran for city council in 2005. Map G shows results for the 2005 general election for Ward 2 between Rapundalo and Republican Thomas Bourque. Rapundalo won the election with 1,574 votes (52%) to 1,436. [Washtenaw County clerk's office results] This map plots Rapundalo’s relative success in the nine precincts of Ward 2. The pushpins indicate where each candidate lives, and the black lines mark the individual precinct boundaries of the ward.

Map G. Rapundalo-Bourque 2005 (Image links to dynamic Google Map)

As in Map F, the shade of green corresponds to Rapundalo’s relative success in a precinct. A darker shade of green means Rapundalo received a higher percentage of votes in that precinct. Rapundalo performed best in precinct 2-2, located in the western corner of the ward, receiving 80% of votes. So 2-2 is the darkest shade of green on the map. He performed the weakest in precincts 2-3 and 2-4, receiving 40.8% of votes. So precincts 2-3 and 2-4 – Bourque’s home precinct and its neighbor to the west – are the lightest shades of all the precincts.

Ward 2: August 2005 – Rapundalo, Kang

Rapundalo contested the August 2005 Democratic primary with Eugene Kang, who was then a University of Michigan student. That was five years after then-Republican Rapundalo contested the 2000 mayoral race against John Hieftje. Rapundalo lost that race.

But Rapundalo won the 2005 Democratic primary race, 575 votes (54.5%) to 480. [Washtenaw County clerk's office results] Map H shows results of that 2005 primary. As in Maps F and G, the pushpins indicate where each candidate lived, the black lines mark the individual precinct boundaries of the ward, and the shade of green corresponds to Rapundalo’s relative success in a precinct.

Map H. Rapundalo-Kang 2005 (Image links to dynamic Google Map)

The darkest precincts are 2-6 and 2-9, where Rapundalo made his strongest showing – 73% and 71% respectively. In the middle three precincts – 2-1, 2-8 and 2-7 – his support was weakest, ranging from 36% to 45%. The southern precincts – 2-3, 2-4 and 2-5 – were fairly evenly split between Kang and Rapundalo and are thus a medium shade of green. In fact the result in 2-4 was a dead heat, with Rapundalo and Kang getting 77 votes apiece.

Precinct 2-2 is completely transparent, because no one voted for Rapundalo there – but that is because no voters showed up in that precinct.

Ward 2: Descriptive Generalizations

While precinct 2-2 was percentage-wise Rapundalo’s strongest in November 2005  and August 2007, in both of those elections, the precinct also had an inordinately low number of total votes – two in August 2007 and fifteen in November 2005. And in August 2005, no one from that precinct showed up to vote. So precinct 2-2 does not factor prominently in the patterns of Ward 2 elections.

The most striking pattern in assessing Rapundalo’s support is how consistently dominant he is in precincts 2-6 and 2-9, which are both in Rapundalo’s immediate neighborhood – Rapundalo lives right on the boundary of the two precincts, in 2-9. In the 2005 primary, Kang was stronger through the middle of the ward, and the candidates were about even across the southern precincts, so those two northern precincts made the difference in the outcome for Rapundalo.

Also in the 2007 and 2005 general elections, those two northern precincts were crucial for Rapundalo. In both of those elections (against Amonsen and Bourque, respectively), if the result had been based just on other precincts, Rapundalo would have lost both elections. Especially in the 2007 race against Amonsen, the southern precinct 2-5 was weak for Rapundalo. That was not just because it is Amonsen’s home precinct. It’s also home to Huron Hills golf course. A perception at the time was that the city was possibly planning to sell the golf course, and that’s what in part prompted Amonsen’s write-in campaign.

It was crucial in both of those elections not just that Rapundalo carry his home precincts, but that he carry them with a high plurality.

Rapundalo’s strength in those precincts is not just a matter of living in the neighborhood. He’s a past president of his homeowners association – the Orchard Hills/Maplewood Homeowners Association – and remains active with the group.

In a 2005 interview, about two months after the election, Bourque reflected on the strength of neighborhood connections and why voters choose one candidate over another:

But they also had to have some kind of personal knowledge or at least try and find out if anybody knew me. I think that’s why Stephen Rapundalo probably won. Because he knew a whole bunch of people in his neighborhood. …  And that’s where, from a vote standpoint, he killed me. In his neighborhood he essentially beat me 2-to-1. And that makes sense. A lot of people knew him and said, I know this guy and I can trust him.

… And you kind of wonder how much of it was the ‘D’ [Democrat] influence and how much was the people he knew. There’s no way to know that. There’s no way we’ll ever know it. Actually other than me and maybe him …. no one will really care!

About the writer: Saline resident Hayley Byrnes is a Chronicle intern. The Chronicle relies in part on regular voluntary subscriptions to support our coverage of elections to public bodies like the Ann Arbor city council. Click this link for details: Subscribe to The Chronicle. And if you’re already supporting us, please encourage your friends, neighbors and colleagues to help support The Chronicle, too!