The North West Angle: how a 1783 screw-up left chunks of US territory marooned in Canada

Spot the deliberate mistake. Image: Google.

Last week I wrote a faintly ranty piece listing 13 things about the map of the US that wind me right up. A lot of people on Twitter pointed out that I'd missed a good one. So...

It's 1783. The American War of Independence is coming to a close, everyone is knackered, and among the many pressing questions facing those drawing up the peace treaty is that of the new nation's northern boundary: where United States territory stops, and British North America begins. This challenge is made all the more difficult by the fact that, west of the Great Lakes, everyone's knowledge of the region's geography gets just a little bit hazy.

But, the negotiators do what they must, and the Treaty of Paris states that the boundary will run:

“...through the Lake of the Woods to the northwestern-most point thereof, and from thence on a due west course to the river Mississippi.”

Back in 1783, the land west of the Mississippi was French territory, so this seemed to settle the matter: find the northwestern-most point of the lake, draw a line west until you hit the river, and then you’re sorted.

How you calculate the northwesternmost point of a lake: it involves drawing a lot of lines from north east to south west. Image: Strafpelaton 2/Wikimedia Commons.

There was just one problem: despite what the maps of that era showed, west of the Lake of the Woods, there was no River Mississippi. The then unknown source of the river was Lake Itasca, in the north of Minnesota about 180km south of the Lake of the Woods.

The red pin marks the source of the Mississippi. Note that it is a surprisingly long way from Canada. Image: Google.

Later treaties would correct this somewhat, by placing the boundary to the west at the 49th parallel (that's the long straight line at the top of maps of the US). To the east, though, it was already established, so the stuff about the northwestern-most point of the Lake in the Woods stayed: the two bits of the border were linked by a straight line running north-south.

The result was a border that looks like this:

Image: Google.

Spot anything odd about that border?

Image: Google.

See it now? Here's another map, from Wikipedia.

Image: The Illusional Ministry/Wikimedia Commons.

That looks weirdly like the top right corner of the map is actually an inset, a larger version of something else on the map. It's not: it's just a part of Canada (Manitoba, to be specific) with square boundaries.

And, because those boundaries are so straight, a couple of bits of land have been cut off, and ended up in the United States.

The larger is the Northwest Angle: home to around 120 Minnesotans, who can't reach the rest of the state without either crossing a lake or going through Canada. It's only 319km2, which isn't very big; but that’s still around twice the size of Washington DC, so it’s not that tiny either.

The other cut off is Elm Point, a small cape which isn't home to anyone at all, but is nonetheless considered US territory. Both of these, incidentally, are “practical exclaves” of the US: portions of territory that can’t be reached by land without going through someone else’s country. (They’re not enclaves because Canada doesn’t completely surround them.)

None of this really matters in practical terms since the US-Canada border is one of the most permeable in the world. There's not talk of building a wall here (wonder why) though you are, technically, meant to phone customs as you cross the border to make your declaration.

But it does teach an important life lesson: don't define boundaries on the basis of a river without checking where that river actually goes first.

For more stories of silly boundaries, click here.

Jonn Elledge is the editor of CityMetric. He is on Twitter, far too much, as @jonnelledge.

Want more of this stuff? Follow CityMetric on Twitter or Facebook.


To transform Australia’s cities, it should scrap its car parks

A Sydney car park from above. Image: Getty.

Parking may seem like a “pedestrian” topic (pun intended). However, parking is of increasing importance in metropolitan areas worldwide. On average, motor vehicles are parked 95 per cent of the time. Yet most transport analysis focuses on vehicles when they are moving.

Substantial amounts of land and buildings are set aside to accommodate “immobile” vehicles. In Australia, Brisbane provides 25,633 parking spaces in the CBD, Sydney 28,939 and Melbourne 41,687. In high-demand areas, car parks can cost far more than the vehicle itself.

However, parking is not just an Australian problem. By some estimates, 30,000 square kilometres of land is devoted to parking in Europe and 27,000 km² in the US. This parking takes up a large part of city space, much of it highly valued, centrally located land.

Traditionally, transport planners believed that generous parking allocations provided substantial benefits to users. In reality, excessive parking is known to adversely affect both transport and land use. These impacts, along with recent land-use, socioeconomic and technological trends, are prompting cities to start asking some important questions about parking.

Australian planners must engage with emerging trends to help cities work out the best way to reclaim and repurpose parking space in ways that enhance efficiency and liveability while minimising disruption.

Here we chart likely challenges and opportunities created by these trends over coming decades.

Key trends affecting parking space in cities. Image: author provided.

Land use

All Australian cities have policies to encourage densification, consolidation and infill development in their centres. In conjunction, some cities are setting maximum limits on parking to prevent it taking over valuable inner-city properties.

Transit-oriented development (TOD) has also become popular, at least on paper. This is another form of urban consolidation around transit nodes and corridors. It is known to benefit from high-quality urban design, “walkability”, “cyclability” and a mix of functions.

These developments mean that people who live in CBDs, inner-ring suburbs and near public transport stops will use cars less. Consequently, demand for parking will decrease.

Some non-TOD suburbs are trying to replicate inner-city features as well. For example, some suburban shopping centres have introduced paid parking. This is a significant shift from previous eras, when malls guaranteed ample free parking.

Suburbanites who lack easy public transport access will continue to rely on cars. But rather than driving all the way to a CBD, commuters will increasingly opt for park-and-ride at suburban stations, thereby increasing demand for park-and-ride lots at public transport interchanges. However, excessive capacity might hurt rather than help patronage.

Social trends

In addition to land use, several social trends will affect the need for parking.

First, young people are delaying getting drivers’ licences because driving is culturally less important to them than in previous generations.

Second, people of all ages are moving from outer suburbs to inner cities. For many, this means less driving because walking, cycling and public transport are more convenient in inner cities.


inally, the emergence of Uber, Lyft and vehicle-sharing arrangements means that people are not buying cars. Research suggests that each car-sharing vehicle removes nine to 13 individually owned vehicles from the road.

Together, these trends point to a reduced need for parking because there will be fewer cars overall.


The importance of technology in parking is rising – paving the way for “smarter” parking.

The emergence of a host of smartphone apps, such as ParkMe, Kerb, ParkHound and ParkWhiz, has begun to reshape the parking landscape. For the first time, users can identify and reserve parking according to price and location before starting their journeys.

Apps also make available a host of car parks that previously went unused – such as spaces in a residential driveway. This is because there was no mechanism for letting people know these were available.

In addition, smart pricing programs, such as SFPark in San Francisco, periodically adjust meter and garage pricing to match demand. This encourages drivers to park in underused areas and garages and reduces demand in overused areas.

The advent of autonomous vehicles promises to have dramatic impacts on transport and land use, including parking.

According to one school of thought, mobility services will own most autonomous vehicles, rather than individuals, due to insurance and liability issues. If this happens, far fewer vehicles and parking spaces will be needed as most will be “in motion” rather than parked most of the time.

More space for people and places

The Tikku (Finnish for ‘stick’), by architect Marco Casagrande, is a house with a footprint of just 2.5x5m, the size of a car parking space. Image: Casagrande Laboratory.

The next decade promises much change as emerging land-use, socioeconomic and technological trends reshape the need for, and use of, parking. Cities will devote less space to parking and more space to people and places.

Parking lanes will likely be repurposed as cycling lanes, shared streets, parklets, community gardens and even housing. Concrete parking lots, and faceless garages will likely be converted to much-needed residential, commercial and light industrial use.

The ConversationBy transforming parking, much urban land can turn from wasteland into vibrant activity space.

Dorina Pojani, Lecturer in Urban Planning, The University of Queensland; Iderlina Mateo-Babiano, Senior Lecturer in Urban Planning, University of Melbourne; Jonathan Corcoran, Professor, School of Earth and Environmental Sciences, The University of Queensland, and Neil Sipe, Professor of Urban and Regional Planning, The University of Queensland

This article was originally published on The Conversation. Read the original article.