Rallies, Rebels and Revolutions

This month we’ll be talking about protest movements in Canada. The inspiration for this month’s topic was the “Freedom Convoy” protests that took place from late January to mid-February. We wanted to see how it compared to other protests in Canada and around the world.

Freedom Convoy Timelines

During the protests it seemed like the crisis was dragging on forever. Some felt that the police were not breaking up the protests quick enough. Also, at the height of the protests, multiple cities were experiencing them at the same time. We wanted to get a better handle on the timelines of the protests and analyze the timeframe with another historical emergency to see how it compared.

First, let’s look at the timelines for the protests themselves:

You’ll notice we only included border blockades and the main Ottawa protest. We had to limit the scope because so many protests happened (and continue to happen) in provincial capitals and other cities across Canada.

As for the comparison to another historical event, we decided to compare it to the 1970 October crisis since that is the only other national emergency where the War Measures Act and its successor, the Emergencies Act, have been invoked. Below, day zero for the October crisis is when the first kidnapping occurred. For the freedom convoy, it’s the first day the protests started in Ottawa:

Note that only major police actions involving arrests of actual FLQ cell members are shown above. There was a large amount of police activity during the October crisis that is not specifically shown.

GiveSendGo Convoy Donations

Doug requested and was provided with access to the leaked data detailing convoy donations to GiveSendGo by the non-profit organization Distributed Denial of Secrets, who publish leaked documents for use by journalists and researchers. Specifically, we used the GiveSendGo 5.0 release.

This data covers worldwide donations to the 2022 Freedom Convoy, but we focused on Canadian donations.

The data takes the form of documents in JavaScript Object Notation (JSON), each detailing a transaction, credit card charge, customer, or event. For our work we focused on the documents detailing credit card charges. Particularly sensitive personally-identifiable information like home addresses and credit card numbers appear to have been stripped from this data but it does provide the date, amount, name and postal code for each individual transaction.

To process this data, Doug wrote a small Python script that processed each charge transaction and produced two output files: a list summarizing aggregate donations by postal code, and another list at the forward sortation area level. You can download these files from this Google Drive folder. 

Combining the FSA-level list with shapefiles at the same level of aggregation allows us to visualize the density of convoy donations across Canada.

Here’s a closeup of Southern Ontario

Western Canada

If you’d like to inspect individual features of the map, you can check out a simplified geoJSON file in the repository.

FSAs are a convenient way to organize data that has postal codes attached to it, but the disparate physical size and population count within FSAs can make the map difficult to interpret. To get around this, we used Microsoft SQL Server spatial tools (as described in this video) to allocate the donations by FSA into federal ridings. Federal ridings are closer to equal in population than FSAs, and offer the additional advantage of being interesting entities in themselves. They still have the disadvantage of varying significantly in size, but we can get around that by visualizing federal ridings as equal-sized squares using the mapcan R library. To build this visualization, we ranked ridings in Canada from 1 to 338 based on the amount of money contributed to the convoy via GiveSendGo and coloured them according to rank. Squares are placed in roughly similar layout to their geographic relationships to one another.

Occupy Movement

The Occupy movement was one of the most memorable protests in recent decades, taking the world by storm in the fall of 2011 and into early 2012 when protests erupted around the world. You might remember most the “We are the 99%” slogan made famous by the movement as participants sought to protest the concentration of wealth among the top 1% of income earners as well as broader social, economic, and political inequality. The protest started as Occupy Wall Street in New York City on September 17, 2011 but quickly spread across 82 countries, with October 15, 2011 marking the biggest global protests with demonstrations in over 900 cities. With no clearly defined goals or structure in the wider movement, and protesters “occupying” parks and other public spaces with campsites for days, weeks, and even months in some cases, this was a unique movement unlike most we’ve seen before. To preface what we found, our data was taken from internet sources and news articles, so there may be some inaccuracies as some movements were covered much more than others, but the below information is accurate according to the available data.

Canada was no exception to the movement’s global spread, and we hosted at least 30 Occupy protests between October 15, 2011 and May 15, 2012. Protests were held in every province, as seen in the map below, made in Excel’s 3D maps tool.

October 15, 2011 was deemed the Global Day of Action, and all 30 Canadian protests began on this day, with some lasting only the day, and some lasting far longer. 

Perhaps not surprisingly, Ontario and British Columbia take the cake for most protests in their provinces, with 9 in Ontario and 8 in BC. Perhaps more surprisingly, given the population of Quebec, is that there were only 2 cities with protests there – Montreal and Quebec City.

We also took a look at the overall length of the movements in each city, as well as whether police force was used in each. 

City	Province	Length in Days	Police Force Used
Calgary	Alberta	57	Yes
Charlottetown	Prince Edward Island	0	No
Comox Valley	British Columbia	0	No
Corner Brook	Newfoundland and Labrador	0	No
Edmonton	Alberta	41	Yes
Guelph	Ontario	29	No
Halifax	Nova Scotia	27	Yes
Hamilton	Ontario	0	No
Kamloops	British Columbia	31	Unsure
Kelowna	British Columbia	0	No
Kingston	Ontario	55	Yes
London	Ontario	25	Yes
Maple Ridge	British Columbia	25	Unsure
Moncton	New Brunswick	138	No
Montreal	Quebec	41	Yes
Nanaimo	British Columbia	55	Yes
Nelson	British Columbia	47	Yes
Ottawa	Ontario	39	Yes
Quebec City	Quebec	38	Yes
Regina	Saskatchewan	33	Yes
Saint John	New Brunswick	31	Unsure
Saskatoon	Saskatchewan	29	Yes
Sault Ste. Marie	Ontario	0	No
St. John’s	Newfoundland and Labrador	213	No
Sudbury	Ontario	10	Unsure
Toronto	Ontario	39	Yes
Vancouver	British Columbia	37	Yes
Victoria	British Columbia	38	Yes
Windsor	Ontario	58	Yes
Winnipeg	Manitoba	66	Yes

On average, movements in Canadian cities lasted 40 days, though this number is slightly skewed by the whopping outliers of Moncton, NB at 138 days and St. John’s, Nfld at 213 days. The median number of days was 35. Why did the St. John’s protests last so long, you might ask? Well, the mayor actively accommodated the protests, unlike in other cities, where action was quickly initiated by city officials to restrict or evict protestors from protest sites. In many cities, the protestors would not freely leave Occupy encampments, and police force was used in at least 17 cities, and perhaps more as information could not be found for all cities.

The 3.5% Rule

The 3.5% rule is based on observation of revolutions and change of government through mass movement or protest. The general theory is that if 3.5% of a country’s population takes to the streets in general protest, this is a critical mass/tipping point where governments tend to fall under that sort of pressure. It has to be noted that these need to be sustained protests and generally they occur in capitals and major cities resulting in the breakdown governing legitimacy. 

This topic has been researched by the Harvard Center for Human Rights Policy and they found that size of protest certainly related to the general success of the movement but with at least 3.5% of the population the success rates approach 90%. There is marked drop off with only 61% being successful when 1-3.5% of the population join in. 

In the report track protests and movements since the mid-1960s to 2014. Generally speaking larger protests tend to be blocked by use of force usually by repressive governments. If a breakthrough occurs it also leads to the downfall of government.

Protest	Year	Estimated Protest Size	Proportion of Population	Notes
2010 Canada anti-prorogation protests	2010	21,000	0.055%
Idle No More	2013	20,000	0.0526%	Estimated from media reports of protests in major Canadian Cities.
Women’s March US	2017	5,000,000	1.50%	Largest single day protest in US History
Floyd George/ BLM Protests	2020	19,000,000	5.72%	Estimated average from across the country at peak of the protest
January 6 Insurrection	2021	2,250	0.00068%	Estimated number who actively entered the Capital
Freedom Convoy – Ottawa	2022	18,000	0.0474%

The above are a few selected examples from more recent history. Anti-Prorogation and Idle No More in 2010 and 2013 respectively garnered only about 20,000 people or 0.05% of the Canadian population. 

The US Women’s March in January of 2017 was the largest single day of protest in US History, with an estimated 5 million people protesting across the country. This number only equated to 1.5% of the US population. 

The BLM/Floyd George protests in the summer of 2020 turned out over 19 million people at their peak which is over 5% of the US population placing it above the threshold for perceived success. In contrast the Capitol Insurrection was perpetrated by only a few thousand people representing 0.0006% of the population.