On the last weekend of May (the first after the death of George Floyd in police custody), I was in San Diego watching local broadcasts of protesters setting fires, smashing windows, looting shops, brandishing weapons, spraying graffiti, and blocking roads in downtown San Diego and nearby La Mesa. At least nine people died that weekend in protests around the US (according to a list on Wikipedia, which is unlikely to over-count). More people died in other areas when law enforcement surged against the protests or were ordered to avoid confrontations. In Chicago, that weekend was the deadliest for homicides (25) since records began.

Yet the online search engines refused to return anything but “peaceful protests.” Siri (the artificially-intelligent friend supplied by Apple) was tweaked to answer the question “Do black lives matter?” with “Yes,” but to answer “Do all lives matter?” with “All lives matter is often used in response to the phrase Black Lives Matter, but it does not represent the same concerns.” In both cases, Siri directs users to the Black Lives Matter website.

Any enforcement of the law gets reported as shocking police violence

The day after that first bloody weekend, Yahoo News admitted no violence in its short report on worldwide “anti-racism protests.” The news agency AFP led with a report  about police and military responses in the US capital, before admitting “isolated looting” in New York, Philadelphia, and Los Angeles. Associated Press reported on foreign protests, with admiration for the toppling of the statue of Edward Colston in Bristol and the protest outside the US Embassy in London (contrary to lockdown). AP admitted that 14 Metropolitan Police officers were injured by “protesters” in one night but claimed these injuries “followed a largely peaceful Black Lives Matter demonstration.”

In the first two weeks after the death of George Floyd, the protests cost $1 billion to $2 billion in insured property damage. In the three calendar months, more than 30 people were killed, and 14,000 people arrested in American protests alone (according to that Wikipedia page). Yet most news media and politicians have followed the line that Black Lives Matter (BLM) protesters are overwhelmingly peaceful. Towards the end of August, CNN infamously broadcast from Kenosha, Wisconsin, in front of burning cars with the banner “fiery but mostly peaceful protests after police shooting.”

The police can’t win. Any enforcement of the law gets reported as shocking police violence. The police withdraw or take a knee. Then the police are blamed for lawlessness. Vigilantes step in. Then the police are accused of colluding with racists.

The narrative goes all the way up to Joe Biden, who excuses BLM’s violence while railing against “right-wing violence” (and editing Donald Trump into the pictures).

Similarly, Britain’s Liberal Democrats adopted the statement: “The Black Lives Matter protests have been overwhelmingly peaceful.” The Liberal Democrats seconded the Labour politicians who accused Boris Johnson of “stoking division and fear” by warning against extremists hijacking minority interests.

The BLM has done well out of this narrative. BLM received millions of dollars in the first half of June alone (including from owners of social media and mainstream media), plus $90 million in funds to bail protesters from jail.

Few newspapers have published scepticism of the narrative. One is the unusually conservative New York Post. An independent American journalist has reported discrepancies between mainstream reporting and his personal observations and interviews in American cities, but his outlets are YouTube and a British online magazine. Most of the criticism of the narrative has come from President Trump and his partisans, whom media like to dismiss (with some justification) as data-free.

In early September, the mainstream media rushed to report supposedly independent findings that 93 percent of BLM protests were peaceful and non-destructive, and that police intervened in more BLM protests than non-BLM protests. CNN claimed that these findings “contradict assumptions and claims by some that protests associated with the Black Lives Matter movement are spawning violence and destruction of property.”

Yet the supposedly independent study is explicitly anti-police, anti-Republican, and pro-protest (of almost any kind). The authors frame their findings to belittle the violence and blame it on counter-protesters. And the 93 percent finding is achieved by under-counting violent protests, using selective sources.

Framing

Let’s start with the framing effects and then we’ll get to the outright under-counting of violence. The self-styled “US Crisis Monitor” counted about 7,750 BLM protests from 26 May to 22 August, of which about 550 were violent. That’s six violent protests a day. Would you be happy to live in a country with six violent protests a day?

The US Crisis Monitor reports “only” 220 locations where protests were violent but does not admit that this works out at a rate of more than 9 percent of the locations in which BLM protests occurred. Imagine you’re a foreign tourist, thinking of visiting 100 American locations. Would you still travel given an expectation that 9 of those visits would expose you to violence?

The report describes 93 percent of protests as “overwhelmingly peaceful,” and describes the other 7 percent as “miniscule.” Imagine if you interacted with 100 protesters, of whom 7 were violent. Would you describe them as miniscule?

Further, the US Crisis Monitor characterises most of the violence as instigated by opponents of BLM masquerading as supporters. However, the report’s few citations focus on a supposed Hell’s Angel and white supremacist wielding an umbrella against windows.

The report then describes the violence as “largely confined to specific blocks, rather than dispersed throughout the city.” What is the threshold for “throughout the city”? The US Crisis Monitor doesn’t explain. It’s a sleight of hand (technically, a fallacy known as incrementalism). This fallacy was mainstreamed earlier, as epitomised by one article’s claim that Portland looks like a war zone only on the internet, and another resident’s declaration that the fires and destruction are “myths” because they apply to certain blocks.

Political Bias

The US Crisis Monitor claims to be non-profit, non-partisan, “independent,” and “objective.” Most of the parent organisation’s funders are foreign ministries, which are interested in the wider programme for collecting data on armed conflict globally. The organisers don’t declare their funders for the American part. They did not return my emails on this question.

The main domestic private funder is the Tableau Foundation, based in Seattle. Its claim to be “data-driven” is undermined by its presumption of “systemic racism” without definition or evidence. Its website claims “the need to dismantle the forms of racisms that are so deeply entrenched in our systems and institutions.” It supplies “data to address … longstanding inequities.” In June 2020 it committed $10 million “to dismantle systemic racism in the US.”

In June, the US Crisis Monitor’s organizers issued a statement expressing “solidarity” with any protests “calling for systemic and peaceful change.” At best, this statement is naive. It literally does not care the aim of the protests, such as authoritarianism or supremacy for one religion, race, or gender. And calling for peaceful change is different to actually protesting peacefully.

The US Crisis Monitor’s report sets up law enforcement as “right-wing,” and aligns itself with the Marxist politics of BLM and Antifa. It complains about “a wider push to militarise the government’s response to domestic unrest, and particularly demonstrations perceived to be linked to left-wing groups like Antifa.” The report repeatedly complains about the Trump administration inflaming tensions, without calling out any opposition politician.

Anti-police bias

The US Crisis Monitor’s report is biased against policing too: “These data reveal that the United States is in crisis. It faces a multitude of concurrent, overlapping risks – from police abuse and racial injustice, to pandemic-related unrest and beyond – all exacerbated by increasing polarization.” Note that the US Crisis Monitor does not admit protesters, or their various enablers, as causes.

According to its report, in about 4 percent of cases the authorities used “force” (such as pepper spray), which the US Crisis Monitor condemns as a “heavy handed police response [that] appears to have inflamed tensions and increased the risk of violent escalation.” This puts the US Crisis Monitor into a ridiculous contradiction: it describes the protesters as overwhelmingly peaceful given 7 percent violence, while condemning the police as heavy-handed given 4 percent.

Biased sources

BLM is replacing law and order with mob rule

The US Crisis Monitor is vague about the sources behind its conclusions. In the dataset, the source for each event is usually a local television station, sometimes a newspaper, sometimes a “local partner.” In no case does the dataset specify the reporter, the date, or any point of contact. In searching for data on my hometown, I found a local television news broadcast in San Diego listed as the source for a protest in Las Vegas (that’s 332 miles away by the shortest road). In 51 cases, the source is simply “Twitter.”

The US Crisis Monitor doesn’t list all its “local partners.” The global research programme talks about “a unique network of international and local partners around the world,” of which it lists a few NGOs, none of them focused on the US. It claims that some sources cannot be disclosed, for “security reasons.”

In the American dataset, some “local partners” have loaded titles, such as “Militia Watch” and “Count Love.” I doubt these partners are doing the police any favours. I didn’t see any partners with names like “Back the Blue.” None of the sources is official.

Coding bias

The US Crisis Monitor doesn’t follow its own coding rules. It defines “violent demonstration” at great length, to include vandalism, looting, blocking roads, and burning tires. By this definition, some cities were experiencing violent protests for months, but that’s not what the US Crisis Monitor admits.

The most observably destabilised city was Seattle, where violence started on 29 May. The Mayor, Jenny Durkan, soon curbed the police response and mischaracterised the protests as a peaceful carnival. The protesters declared a police-free zone on 8 June, initially under the title of CHAZ (Capitol Hill Autonomous Zone), later as CHOP (Capitol Hill Occupied Protest). Predictably, the area became a haven for crime. The Mayor’s own office admitted at least five times more criminal incidents within the area in June 2020 compared to June 2019, including two excess murders and 16 excess aggravated assaults. (The reports were gathered outside CHOP, so the true rate must have been higher.) On 1 July, city and federal law enforcers cleared the streets, but crime remained unusually high through August. City politics continue to favour the protesters. For instance, city government recently closed a park against a Christian gathering, but opened other parks to protests against the police.

Counting from 29 May to 1 July (i.e., ignoring disturbances after CHOP was cleared) gives 34 days. By the coding rules, every one of those days should count as violent, just for the disruption to traffic. Yet the dataset contains only 28 non-peaceful protests in Seattle from 26 May through 22 August (some of them occurring on the same day), plus 59 peaceful protests. The qualitative report does not give these findings but condemns police response.

The US Crisis Monitor is pretending to be data-driven where they are agenda-driven

The report has more to say about Portland, which suffered violence and destruction for 47 days in a row, according to the Department for Homeland Security, from 29 May until mid-July, when city and federal officers shut down the short-lived Chinook Land Autonomous Territory (CLAT). The US Crisis Monitor’s dataset shows protests on every subsequent day through 22 August. Indeed, only four of the 89 days from 26 May lack a protest in Portland, according to the dataset. Usually, multiple protests are recorded in a day. In total, the dataset shows 159 protests in Portland over those 89 days, of which 83 protests (52 percent) were other than peaceful. For whatever reason, the coders in Portland are more honest than those in Seattle.

Yet the US Crisis Monitor does not report these disturbing findings. Instead, it blames federal intervention for a jump in violent protests in the city and in the rest of the State of Oregon. Why is that the fault of the law enforcers rather than the protesters?

The US Crisis Monitor, its sponsors, and its reporters are pretending to be data-driven where they are agenda-driven. The project is anti-police and pro-Marxist. Its data collection does not even conform to the project’s own coding rules. Its sources are biased.

BLM betrays its own anti-empiricism with memes such as “silence is violence” and “all whites are racists.” In practice, whatever its virtuous intentions, BLM is replacing one racism with another, law and order with mob rule, discourse with confrontation, democracy with a tyranny of the violent minority.

The road to hell is paved with good intentions – and bad investigations.