Overall verdict: Partly accurate, but ideologically overstated

The article contains several genuine and important research findings. However, it selectively interprets them to support a political conclusion that the studies themselves do not necessarily establish.

Its strongest argument is that poverty does not automatically turn people into criminals, and individual responsibility cannot be ignored.

Its weakest argument is that because some studies find no independent effect of family income, poverty and social conditions are largely irrelevant to crime.

Let’s examine the evidence.

1. Swedish sibling study – the sevenfold difference

The article attributes this finding to Sariaslan et al., British Journal of Psychiatry, 2014.

The underlying argument is that children from poorer families have substantially higher rates of violent crime, but comparing siblings reduces or eliminates the association.

There is an important scientific distinction here:

• A population-level association measures differences between children from different families.
• A sibling comparison controls for many characteristics shared by siblings, including much of their family background.
• It does not control for everything that differs between siblings.

A sibling study can therefore strengthen causal inference, but it is not equivalent to a randomized experiment.

Verdict: The general interpretation is plausible, but the article needs to provide the exact Swedish paper and its statistical results. I would not accept the stated sevenfold figure and hazard ratio of 0.95 as verified from the citation as presented.

2. Finnish study – 650,680 individuals

Substantially verified

This is the strongest piece of evidence in the article.

The actual study was published in the International Journal of Epidemiology in October 2021.

Authors examined 650,680 individuals born between 1986 and 1996, including 426,886 siblings.

The findings were: Analysis Association with violent crime arrest Population-level analysis HR 0.77 per $15,000 higher income Sibling comparison HR approximately 0.99–1.00

In other words, the association between higher family income and lower violent-crime arrests disappeared in the sibling comparison.

Oxford Academic +1

But there is an important qualification.

The study investigated childhood family income and subsequent outcomes, including violent-crime arrests. It did not establish that poverty has no influence on every type of crime, nor that all poverty-reduction policies are ineffective.

It also cannot rule out effects of neighbourhood deprivation, unemployment, social exclusion or other economic circumstances.

My interpretation: this is strong evidence against treating family income alone as a direct, independent cause of violent offending in this particular population.

3. The 2008 financial crisis – crime went down

The US statistics are correct

The FBI’s 2010 crime statistics confirm:

• Violent crime fell by 6% compared with 2009.
• Robbery fell by 10%.
• Murder fell by 4.2%.
• Property crime fell by 2.7%.

Table 1 +1

This challenges the simplistic proposition that increasing unemployment must immediately increase crime.

However, there are several problems with the article’s conclusion.

First, the recession was not a controlled experiment. Crime is affected by numerous variables, including demographics, policing, drug markets, incarceration, opportunity and social behaviour.

Second, the relationship between economic hardship and crime may operate over longer periods.

Third, not all crimes respond equally to economic conditions.

Verdict: Correct observation, incorrect causal conclusion. Falling crime during a recession does not prove that poverty has no effect on crime.

4. Moving to Opportunity – the randomized experiment

The article misrepresents the findings

This is particularly interesting because it was a randomized experiment, not simply an observational study.

The US government offered housing vouchers to families living in high-poverty public housing areas. Some vouchers were restricted to low-poverty neighbourhoods; other families received standard vouchers, while a control group received no voucher.

The study examined 4,643 young people.

The results were:

• Boys in the experimental group initially experienced substantial reductions in violent-crime arrests.
• Property-crime arrests increased.
• The effects diminished over time.
• Neighbourhood improvements also diminished over time.

The researchers concluded that current neighbourhood conditions appeared more important than past neighbourhood conditions in explaining criminal behaviour. PubMed +1

Why this matters: The experiment actually provides evidence that changing the environment can change criminal behaviour, at least temporarily.

The statement “New zip code. Same behaviour” is therefore misleading.

5. Police effectiveness – the Chalfin study

Correctly cited

The 2022 American Economic Review: Insights study by Chalfin and colleagues found that:

• Each additional police officer was associated with approximately 0.1 fewer homicides.
• The per-capita reduction was twice as large for Black victims as for White victims.
• Larger police forces also increased arrests for certain low-level offences, disproportionately affecting Black Americans.

American Economic Association +1

The article correctly highlights the protective role of policing.

However, there is a distinction between supporting effective policing and advocating indiscriminately for more policing.

Good policing requires:

• Adequate police presence.
• Prompt investigation.
• Fair and impartial enforcement.
• Accountability against abuse.
• Public trust and cooperation.

The study does not establish that every increase in police manpower, regardless of deployment or quality, will produce the same benefit.

6. Dunedin study – self-control and criminal behaviour

Research accurately represented

Moffitt and colleagues published A Gradient of Childhood Self-Control Predicts Health, Wealth, and Public Safety in PNAS in 2011.

They followed 1,037 children from birth to age 32.

Self-control was assessed repeatedly between ages 3 and 11, using observations and reports from parents, teachers and children.

The researchers found that poor childhood self-control predicted adult criminal convictions, even after accounting for childhood socioeconomic status and IQ.

The association was also examined in a separate sample of 509 twin pairs. moffittcaspi.trinity.duke.edu +1

An especially important finding was that individuals whose self-control improved as they grew older had better adult outcomes, even after accounting for their initial self-control.

But the authors explicitly acknowledged that this was not an experimental intervention. Natural improvement does not automatically prove that a particular intervention will produce the same results. moffittcaspi.trinity.duke.edu

My interpretation: Personal behaviour matters, but behaviour is also capable of improvement. This supports early childhood development, parenting interventions and education rather than simply punishment.

7. Swedish cognitive ability study – 700,514 men

Partly accurate, but selectively presented

The 2012 study by Frisell and colleagues examined 700,514 Swedish men.

It found an inverse association between cognitive ability measured at age 18 and subsequent violent-crime convictions.

The association remained after adjustment for childhood socioeconomic factors.

However, sibling comparisons reduced the association further.

For example, the coefficient fell from approximately −0.18 after socioeconomic adjustment to −0.10 in full-brother comparisons. PubMed +1

Thus, the article is right that cognitive ability predicts violent criminality independently of measured socioeconomic status.

But it is wrong to imply that the relationship is almost entirely unaffected by family background.

There is another important caution: cognitive ability is a statistical predictor, not a moral characteristic. It should never be used to label individuals as potential criminals.

8. Great Depression – the 25% unemployment argument

Insufficiently substantiated

The article claims that crime declined in many American cities despite unemployment reaching approximately 25%.

This is a frequently repeated historical observation, but it needs much more precise sourcing.

There was no modern, nationally consistent crime-reporting system comparable to today’s FBI Uniform Crime Reporting system throughout the Great Depression.

Crime statistics from individual cities, changes in reporting practices and differences between crime categories make sweeping comparisons difficult.

I would not use this example as decisive scientific evidence.

Overall assessment of the article

Claim Fact-check Swedish sibling study Finnish 650,000-person study Crime fell during the recession Recession disproves poverty-crime link Moving to Opportunity showed no behavioural change Police presence reduces homicide Self-control predicts criminal convictions Improved self-control predicts better outcomes Cognitive ability predicts violent crime Great Depression proves poverty irrelevant

9. What does the broader scientific evidence suggest?

There is a distinction between three propositions:

• Poverty directly causes criminal behaviour.
• Poverty increases exposure to circumstances that can increase the risk of criminal behaviour.
• Poverty makes criminal behaviour inevitable.

The first is too simplistic. The third is plainly false.

The second is consistent with a much broader understanding of criminology.

For example, poverty may be associated with:

• Greater exposure to violence.
• Poorer educational opportunities.
• Family stress and instability.
• Concentrated neighbourhood disadvantage.
• Reduced legitimate employment opportunities.
• Greater exposure to criminal networks.

These are potential pathways, not inevitable consequences.

And there is an important distinction between absolute poverty, relative deprivation, unemployment and neighbourhood disadvantage. They should not be treated as interchangeable variables.

10. My conclusion: Responsibility and social justice are not opposites

I think the original article makes an important moral point but draws an unnecessarily ideological conclusion.

Consider two individuals:

• One grows up in poverty, struggles through life, raises a family and never commits a crime.
• Another grows up in similar circumstances but chooses to rob, assault or exploit others.

We should not insult the first person by suggesting that poverty makes criminality inevitable.

Nor should we excuse the second person’s actions merely because of his circumstances.

But neither should society ignore conditions that increase the risk of crime.

A sensible criminal-justice policy needs two parallel approaches:

1. Prevention

Quality education, early childhood development, employment opportunities, family support, safer neighbourhoods and rehabilitation.

2. Protection and accountability

Effective policing, timely trials, proportionate punishment, victim protection and equal application of the law.

Neither approach can substitute for the other.

A particularly important Indian perspective

In India, this debate becomes even more relevant.

A poor person living in a village, slum or economically deprived urban neighbourhood may have limited access to education, employment, healthcare and legal protection. Yet millions of people living in precisely these circumstances remain law-abiding.

At the same time, serious economic offences, corruption, fraud and exploitation are also committed by wealthy and educated individuals.

Therefore, crime cannot be understood simply through income.

The justice system must be equally accessible to the poor and the rich, and the law must be equally applicable to both.

A scientifically stronger version of the original statement

I would replace its opening sentence with this:

Poverty does not create criminals, and wealth does not guarantee morality. Criminal behaviour is influenced by a complex interaction of individual responsibility, family environment, childhood development, social circumstances and the certainty of law enforcement. A society must address the conditions that encourage crime without excusing those who commit it. Compassion for the disadvantaged and justice for the victim are not competing values.

My final verdict: The article is a useful challenge to the simplistic slogan that eliminating poverty will eliminate crime. But it goes too far in suggesting that poverty and social conditions are largely irrelevant. The scientific evidence supports neither extreme.

The real objective should be a society that reduces the causes and opportunities for crime, protects innocent citizens and holds offenders accountable, irrespective of their economic or social position.