The Rise of Vibes as Political Truth

Is free speech at risk? Outsourcing our opinions puts us at the mercy of master manipulators. It’s time to take The Sniff Test.

At the Mercy of Vibes

The headline claims London’s Low Emission Zone (LEZ) improves students’ test scores. The media responds with vibes. These are jumps from fact to fiction without consideration of the evidence.

A study that should generate debate is read one of two ways. The first is that all traffic schemes are great. The second is that academic research is biased.

Free speech enables all views to be heard. It provides protection against tyranny and the crushing of minorities by majority will. The legal basis is the rights of nobles to petition the King under Magna Carta.

The First Amendment to the US Constitution upholds the right to free speech, as did Article 11 of the revolutionary French Declaration of the Rights of Man. This however, is qualified in a way that remains important in European politics today.

“The unrestrained communication of thoughts and opinions being one of the most precious rights of man, every citizen may speak, write, and publish freely, provided he is responsible for the abuse of this liberty, in cases determined by law.”

Whoever controls the law dictates what might damage society. When the law bends to populist sentiment rather than principle and precedent, individual rights are at risk. The UK government is accused of two tier policing and European leaders threaten US tech titans for broadcasting on their own media.

Yet even the degree to which this is happening is questionable. A search for the number of people arrested for speech crimes links back to a 5-year old unsubstantiated claim that the UK arrested more people than Russia. One dubious claim that goes viral can anchor the debate in a way that’s hard to shake off.

This is where academics fill the void. By finding and analysing data, researchers get us nearer to the truth. How should we interpret their findings? If we cannot answer this question, then we are at the mercy of vibes.

Pollution is Bad

Pollution is bad. The clue is in the name. It is not controversial to believe that lower road pollution will have health benefits. Proving this is hard however, and is not proof that the method of lowering pollution is the best one.

Western societies have a poor record of drawing lines on a map to determine political outcomes. Africa and the Middle East are divided in less than ideal ways. But we have yet to come up with a better way of creating political units.

The London LEZ is lines on a map. It was introduced in 2008 and covers most of Greater London. It was complemented by the Ultra Low Emission Zone in 2019, which is now extended to the city’s 580 square miles and 9 million people.

I walk much more than I drive in London. I prefer clean air to dirty after my experience in Hong Kong, peeling pollution from my face with blackened finger nails. I believe that progress means freedom of movement with lower adverse health effects.

I also favour local communities. I believe they can be fantastic support networks that counter isolation and depression. Therefore, I am predisposed to oppose any restrictions on who has access to them. This makes me suspicious of traffic schemes.

My first response to the idea that the LEZ boosted test scores was sceptical. Have people not heard of wind? An MP claimed that only half her constituents had clean air because the ULEZ boundary split her territory in two. This is a text book vibe, transforming the fact of ULEZ, into the fiction that lines on a map control the spread of pollution.

There was no escape. I was going to have to read the LSE paper.

Freethinking Freaks

Between 1991 and 2001, violent crime in the US fell more than 30%. The only comparable period was after the repeal of Prohibition. This seismic change required an explanation.

In 2005, Steven Levitt and Stephen Dubner published Freakonomics. It is a book about rigorous data analysis explaining real world events. It is both controversial and a best seller and remains one of my favourite reads.

The book’s origins were a paper Levitt wrote in 2001 exploring the reduction in crime rate. He ruled out economic growth, an ageing population, gun-control laws, right to carry laws, and capital punishment as reasons. He attributed a modest benefit from increased policing, more imprisonment, and the ebb and flow of the crack epidemic. But there was a bigger factor as yet unidentified.

Levitt is a lateral thinker. New knowledge comes from unconventional thought, forming a theory and testing it. Levitt figured that legalised abortion after 1973 might have had an effect. After all, the reduction in crime was at the time when the never-born children would have been coming of age.

The data analysis confirmed Levitt’s thinking. It also produced a political backlash from left and right. This response is a common one, whereby we dismiss evidence because it does not confirm our beliefs.

On the left, people were outraged at the suggestion that poor people and those from broken homes commit crimes. There was also the hint of eugenics. That’s the deliberate breeding of people on the assumption that politicians are better than nature at deciding who should prevail.

Levitt rubbished this claim, noting women were best placed to know whether they could support bringing a child into the world. That was nothing to do with selective breeding.

Attacks from the right focused on abortion. If it reduced crime, then it was good. That’s bad for those opposed to abortion on religious grounds. Media coverage turned the analysis into an issue of race, even though that was not mentioned in any explanations. Racism was the assumption of those reading, or more likely hearing about the analysis.

I noted last week that economic models are simplifications of the world. As such, they cannot explain its complexities, although the better ones provide an approximation. Levitt reviewed ten potential factors that could have caused crime rates to fall. Four had an effect and abortion was by the far the most significant. But he could not test what he hadn’t considered.

Jessica Wolpaw Reyes is a professor of economics obsessed by the impact of environmental toxicants on social behaviour. To her, there was a glaring omission in Levitt’s work. The Clean Air Act of the early 1970s coincided with the legal right to abortion.

Lead is associated with reductions in I.Q., deteriorating social behaviour, and heart and kidney ailments. It was prevalent in petrol and paint and they were everywhere. This was about to change.

Reyes ran an analysis and concluded that both abortion and the reduction in use of lead were significant in reducing crime in America. Levitt agreed, saying,

“Look, the world is complex and there could be many things going on.”

In academia this is caused multivariate causality. The vast majority of statistical papers attempt to control for it and, as Levitt did, eliminate the possible causes that the data does not support. But what is left at the end of a statistical analysis is not definitive proof of causes. It is simply the best explanation of the authors of the analysis.

This does not go down well in politics and the media. They demand answers. If there is a problem, then there must be a solution. This has two consequences.

The first is hijacking academic work to make political points. Politicians are bigger mouth pieces than professors and can make exaggerated claims stick. Researchers can also be influenced to come up with conclusions. If you are a big company or political donor that funds research, you can have an outsized impact on the conclusions through choice of who makes them.

Levitt’s solution for this is to discuss the data and not the hypothesis. Intelligent people can be presented with a series of facts about the world and left to debate what they mean. This is the opposite of what happens today, when a theory is accepted or rejected regardless of evidence and based on preconceived political opinions.

Assimilation and Clean Air

It’s widely accepted that air pollution is a problem. The World  Health Organisation says it caused 4.2 million premature deaths in 2019 and that 99% of the global population live in areas below acceptable levels. Anyone walking besides a main road understands that traffic is a significant cause of pollution. It is also a necessity.

The LSE paper compares schools inside the London LEZ with 20 English cities more than 100 km away. It analyses Key Stage 2 test results at age 11. These include standardised tests in English and Maths with external examiners.

Source: Avila-Uribe et al.

The improving test scores in London compared to the rest of England are statistically significant. This means they are not caused by chance, but it does not necessarily mean they are caused by what is being tested. Hence the authors adjust for some other factors that might trigger better results.

I remember reading a study into the most effective way to boost educational standards in poor countries. The winning solution is not better teaching, more books or writing materials, but feeding the kids. When they aren’t hungry, they concentrate on lessons.

The LSE authors conclude that lower pollution is at least as effective as raising the quality of teaching, reducing class sizes by 10 pupils, and paying financial bonuses to teachers. They also conclude that the benefits of low pollution increase through time and that children from low income families improve the most.

One limitation of the analysis is that it runs from 2005-15. Key Stage 2 results across the country trended up from 2005 to 2019 and while the analysis adjusts for this general trend, there is a bigger issue. There was no testing during Covid and results since have not attained pre-pandemic levels. Shutting schools and remote teaching have lasting impacts on performance that outweigh air quality.

London schools also contain a lower number of native English speakers. Rapid progress in reading and writing might be due to a greater share of pupils recently learning the language. This would be a victory for assimilating immigrant communities rather than traffic controls.

The Pupil Premium, introduced in 2011-12, provides additional funding for low income students. Even after adjusting for this, the paper claims that pupils from lower socio-economic backgrounds benefit most the LEZ. Low income schools are defined as those with more pupils receiving free school meals.

We already know that provision of food is one of the most effective ways of improving test scores. I would expect school performance to improve the longer children were fed. Remember that all the paper shows is that London gets progressively better than the rest of the country from 2005-15.

Do we have an explanation why disadvantaged children do better in the LEZ? Are poor people more affected by air pollution? If so, then reducing it is an egalitarian measure. If not, then one of the main conclusions of the paper is thrown into contention. We might assume poor people live nearer busy roads, even in central London, but there is no evidence presented either way.

A Little Knowledge

The paper is sound as far as it goes. Pollution is harmful and less of it is a good thing. That is not the same as saying Low Emission Zones are the best way to reduce pollution, or an effective way to boost school grades. For example, vehicle owners can pay to drive high polluting vehicles in restricted zones. Others retrofit vehicles, which limits pollution from fuel, but not from tyres or damaged streets.

The paper’s authors present it as the start of a discussion. Someone must replicate its findings. It will be hard to update them given the introduction of ULEZ and the creation from 2021 of similar zones in several major UK cities.

As with a lot of academic work, the issue is not the limited scope of the research, but the broad application of it for political ends, both for and against. In a world where the first word is the accepted word, with corrections and qualifications buried and forgotten, a little knowledge is a dangerous thing.

Levitt’s proposal is a guide for all of us. Examine the data and draw our own conclusions. The more we rely on media interpretation, the more we are manipulated by vibes.