Maps elsewhere: Mapping the leaked BNP membership list

A few weeks ago, a disgruntled former member of the far-right British National Party leaked the party's membership list. The Guardian explains:

Every police force in the UK was tonight scouring the leaked British National party membership list for names of serving officers, after the Merseyside force confirmed it was investigating one officer's links to the far-right party.

The Prison Service pledged to oust any employee on the list and far-right supporters spoke of fear for their livelihoods as the BNP was plunged into crisis.

...Around 13,000 names and addresses appear on the list. It is thought to include a number of lapsed members and possibly the names of individuals who contacted the party and expressed interest in joining but did not do so.

However, the BNP has confirmed that it is essentially a genuine membership list. It includes the names of a number of clergymen, an actor, two solicitors, at least one doctor and a number of primary and secondary school teachers.

Full story

Both the Guardian and the BBC's Map of the Week have presented the data as a map. They took different approaches:

The Guardian's map (link) mapped the list by constituency using Flash, with a limited number of areas clickable. It's more or less impossible to get a sense of the underlying geography. Also, limiting the number of clickable polygons means some questions aren't answered. What is the constituency between Glasgow and Edinburgh which has the highest BNP membership in Scotland? There's no way of finding out.

The BBC's map (link) mapped the list by postal area, presenting the map as a 432x500 .jpg covering all of Britain, which limits the reader to noticing general regional patterns:

...There is a spine of BNP support running down the Pennines from the former Northern English mill towns of Blackburn, Burnley and Bradford to Derby, Nottingham and Leicester in the East Midlands.

Their choice of colours is more intutive than the Guardian's map, to my eye.

However, both maps use the absolute number of party members in any given area, while I think it would be more meaningful to look at membership per 1,000 of population. This is particularly important when using postal areas where (unlike electoral divisions) there is no real attempt made to make them equal in population.

Map of the Week

will return next week.

Map of the Week: Impaired driving, part 1 (updated)

CARLOS OSORIO/TORONTO STAR

This week starts our two-part series on impaired driving.

The map shows the top 20 of Ontario's 523 postal areas for driver's licence suspension for impaired driving in 2007. I excluded two low-population areas.*

There are several distinct clusters:

Northwestern Ontario stands out, with Sioux Lookout (#1 on the list), Kenora and Fort Frances all in the top 20.

Communities in the Muskoka-Parry Sound area have four places in the top 10.

Two postal areas in and around Brockville are in the top 20.

Both of Cornwall's two postal codes are on the list.

Preston and a nearby Cambridge postal code are #20 and #19, respectively.

The information was obtained from the Ministry of Transportation under freedom-of-information legislation. It is taken from the database that supports this program (scroll down to Immediate Licence Suspension).

Using the FOI process sometimes seems to imply some tension between a journalist and a branch of government. That is certainly sometimes the case. However, in this instance, the MOT very helpfully threw in the age and sex of the suspended drivers, data I hadn’t thought to ask for.

Here is what we learn from this extra information:

The male-female ratio is about 7:1 (14,551 to 2,365).

The age graph is surprising. There are very few teenagers. The graph rises steeply at 20, peaks at 23, declines gently after that point, then starts to rise again at 40, hitting a second peak at 45. After that point, it shows a clear second decline.

---

Click on the chart to see the full image.

---

Update: It occured to me to create the age graph for female drunk drivers (who, it bears repeating, are a small minority at just under 14% of the total). The age graph for male drunk drivers is perhaps not surprisingly similar to the one above. The graph for women, on the other hand, has some differences:

---

Click on the chart to see the full image.

---

Women drunk drivers turn out to be more middle-aged than their male counterparts. While the highest age group for male drunk drivers is 23, the highest age group for female drunk drivers is 43.

---

In part 2, which may or may not run next week, we will look at the impaired driving map of the GTA.

Map may display better in Firefox.

* L0H (Gormley), with five suspensions and a population of 713, and K1P (central Ottawa) with two suspensions and a population of 396.

Thanks again to Star computer-assisted reporting editor Andy Bailey for his help with the data.

Map of the Week: Lack of English fluency

The 2006 census found that English fluency* falls below 90% in two areas of the GTA (map):

1)  the neighbourhoods immediately east and west of downtown (three postal codes centring on Dundas and Ossington, Dundas and Spadina and Queen and Carlaw), and

2) northwestern Scarborough and a pocket of Markham (two neighbourhoods centred on Finch and McCowan, another to the west centred on Finch and Warden, and another in Markham centred on Finch and Hwy. 407.)

* (total census respondents minus those saying they spoke French only or neither English nor French)

Map of the Week: Crime

MICHAEL STUPARYK/TORONTO STAR

This week’s maps depict Toronto crime rates by police division, drawn from from this page.

The six maps cover sexual assault, assault, robbery, breaking and entering, auto theft and theft over $5,000.

The available crime data covers the period from January 1 through August 28 of this year, so the crime rates per 1,000 should not be read as annual crime rates. Its usefulness is showing how crime, or at least reported crime, varies between neighbourhoods.

Downtown’s 52 Division (which has a resident population of about 41,000, far lower than the average of about 161,000) is included on the map out of consistency, but this is one case where looking at a ratio of residents and reported crime breaks down. There's disproportionately people and activity compared to the resident population.

There are certain patterns. Breaking and entering is concentrated in 55, 51, 52 and 14 Divisions, in the central city. The city's northwest (23 and 31 Divisions), on the other hand, seems to suffer more car theft.

The Toronto police publish this disclaimer about the data.

Maps elsewhere: The U.S. election

Google's LatLong blog linked recently to an outstanding U.S. election map working down to the county level (select a state to show county maps) which uses the Google Charts API in much the same way we do, but has several bells and whistles that I've never seen before: very fast loading, a mouseover feature on the state and county polygons, and an option to embed the map, as seen below: