I think most immigrants become either extreme winners, extreme losers/going on benefit or become extreme hard workers. The situation of immigrants leave few medium choices.
1. Start with a dumbed-down version of an opposing position: immigration causes crime.
2. Refute it using propositions that few serious opponents would dispute: immigrants are heterogeneous; socioeconomic conditions matter; migration can have benefits; crime has multiple causes.
3. Present that refutation as though it settles the substantially harder disputes people are actually having: the effects of the scale and composition of migration, integration outcomes, specific forms of crime, fiscal effects, segregation, and the effectiveness of particular policies.
Britain provides an even more amusing historical experiment.
Between 1788 and 1868 Britain and Ireland transported about 160,000 convicted people to Australia. Many had been convicted of theft and other property offences.
Crime remained a major social problem in nineteenth-century Britain because the conditions that generated theft, violence and disorder remained: urban poverty, insecure employment, overcrowding, extreme inequality, alcohol problems and enormous social disruption during industrialisation.
That example reveals the central error in treating crime as something residing inside particular populations.
You can move people.
Unless you change the conditions, other people can occupy the same social position.
This does not mean that poverty mechanically turns people into criminals. Most poor people never commit serious crimes. Nor does inequality excuse a robbery or shooting. Explanation and moral justification are completely different things.
But criminology asks why the probability of offending differs between environments.
Here inequality, deprivation, social stress, segregation, criminal opportunity, weak social trust and access to legitimate employment repeatedly matter.
Immigration can interact with these variables. A very rapid refugee inflow can strain housing, schools and local services if government refuses to expand capacity. Poor integration policies can create unemployment and segregation. Particular criminal networks can develop within particular migrant communities.
Migration is one possible input into that system. It is not the system itself.
So the opponent's third point does not rescue his argument.
Yes, investigate scale.
Yes, investigate composition.
Yes, investigate integration.
Yes, investigate particular crimes.
Yes, investigate fiscal effects.
But those are empirical questions, not evidence that immigration causes crime.
And fiscal effects do not logically prove a criminological effect at all.
The scientifically defensible conclusion is therefore more modest than “inequality explains all crime”, but considerably stronger than the anti-immigration position:
Immigration is not a general or sufficient explanation of crime. Socioeconomic deprivation, inequality, segregation, social stress, criminal opportunity and institutional conditions have much broader historical and empirical explanatory power. Migration sometimes affects crime because it interacts with those conditions. Good integration policy can therefore reduce the risk, while austerity, exclusion and concentrated poverty can increase it.
The criticism sounds rigorous, but it does not actually refute my argument.
It is true that serious research must examine the scale and composition of migration, integration, segregation and different types of crime. But simply listing those variables does not show that immigration is the cause of crime. They have to be tested empirically.
And when researchers do that, the picture becomes much more complicated than “more immigration = more crime”.
Sweden's National Council for Crime Prevention, Brå, reviewed Swedish and international research in 2025. The report was produced following a government assignment and therefore emerged in a political context in which the Tidö government had explicitly made reducing immigration a central objective. That is worth mentioning. But there is no evidence that the government ordered Brå to reach a predetermined conclusion.
More importantly, Brå did not establish such a conclusion anyway.
Brå writes that there is no simple relationship between migration and national crime trends. International research shows different results for different immigrant groups, countries, periods and crimes. Some studies find increases in particular offences; others find no significant effect, and some find lower crime associated with immigration.
Brå also stresses that the effects of migration are small compared with all the other factors affecting crime in both migrant and native populations.
That is already enough to expose the logical problem in the original criticism.
Nobody disputes that the composition of a population matters. A population containing many young men will normally have more crime than an older population, regardless of ethnicity. Refugees arriving after war may initially have worse labour-market conditions than recruited engineers. Migrants placed in segregated neighbourhoods with poor schools and unemployment face different conditions from migrants entering employment immediately.
But these observations weaken rather than strengthen a simple ethnic explanation.
They tell us that conditions matter.
The same principle becomes particularly obvious when we examine native-born Swedes.
Sverigedemokraterna is a useful control case. I am not claiming that being a Sweden Democrat causes criminality. That would reproduce exactly the statistical error I am criticising.
But SD candidates have repeatedly been unusually overrepresented in Swedish investigations of criminal records.
In a 2018 review, more than one in twenty SD candidates had been prosecuted for crime during the period examined, the highest share among the parties. Another local SVT investigation in Örebro County found that roughly one in five SD candidates had criminal convictions; offences in the candidate population included aggravated assault, narcotics offences and attempted arson.
The pattern did not disappear afterwards. Acta Publica found 64 candidates from parliamentary parties with links to criminal motorcycle environments before the 2022 election. It also identified 289 candidates who had expressed Nazi or racist ideas or had similar far-right connections; 214 of those candidates represented SD.
In 2026, another large investigation found criminal convictions during the previous ten years among 3.3 per cent of SD candidates, compared with 1.1 per cent among Social Democratic, Green and Centre Party candidates.
Again, the correct conclusion is not that Sweden Democrats are inherently criminal.
The interesting conclusion is the opposite: statistical overrepresentation does not explain its own cause.
We therefore have to ask why particular groups recruit or contain more people with certain social backgrounds.
That principle must apply equally to immigrants.
The history of SD itself also makes simplistic talk about a uniquely immigrant culture of violence rather awkward. During the 2010 incident later known as the “iron-pipe scandal”, leading SD representatives Erik Almqvist, Kent Ekeroth and Christian Westling became involved in an aggressive confrontation in Stockholm. Video showed racist and sexist abuse, and Almqvist and others armed themselves with metal pipes.
More recently, violent far-right networks known as Aktivklubbar have grown in Sweden. These are not representative of ordinary conservative voters and should not be confused with the democratic right. They belong to the neo-Nazi and militant far-right environment. In 2025 four members of Aktivklubb Sverige received prison sentences after unprovoked attacks on people selected because of their foreign background.
There is an additional irony. A leading member of the criminal motorcycle gang Comanches attended Jimmie Åkesson's wedding in 2024. Åkesson said afterwards that he had not known about the man's gang connections. That qualification matters. But Swedish police nevertheless described the relationship as problematic because organised motorcycle gangs actively seek political influence.
None of this proves anything biological or cultural about right-wing Swedes. Precisely.
It shows why we should not make that inference about immigrants either.
There is much stronger evidence for socioeconomic mechanisms.
Wilkinson and Pickett's The Spirit Level became famous for showing that more unequal affluent societies tend to experience worse social outcomes, including more violence. Their original work has attracted legitimate methodological criticism, so it should not be treated as the final word. But subsequent research has repeatedly found an association between income inequality, low social trust and homicide.
One cross-national study of 33 countries found homicide and assault to be more common where income inequality was greater and identified social trust as an important mediating mechanism.
A 2024 re-examination of the Spirit Level thesis using updated data and additional controls also found substantial continuing evidence connecting inequality with social harm, although individual outcomes vary.
Researchers studying Britain's Welfare Reform Act of 2012 compared places differently exposed to welfare cuts. Areas hit harder by the cuts subsequently experienced larger increases in crime, driven especially by violent crime. The effect was strongest in neighbourhoods that were already deprived.
That is much closer to a natural experiment than simply correlating ethnicity with offending.
History also destroys the idea that modern immigration is necessary to produce high levels of violence.
Swedish historians estimate that homicide rates in the fifteenth and sixteenth centuries may have been around twenty times modern Swedish levels. Sweden obviously did not have today's refugee immigration, multicultural suburbs or Middle Eastern gangs.
European homicide subsequently declined enormously over several centuries. That decline began before the modern welfare state, so it would be wrong to credit the welfare state with all of it. Historians point to stronger state institutions, declining private violence and changing social norms among the causes.
But the historical evidence establishes something simpler and very important: large-scale violence existed perfectly well without modern immigration.
It may be true that non-European immigrants do best in higher education, but they also do "best" in crime. I wonder why. Are there two distinct groups going two different roads?
A private theory is that those who come for working do better, and those who flee from violence do worse. Those who come for work take a step to a better life, those who flee from violence may end up in desperation.
I would like the theory to be tested. Or if that has already been done I would like someone coming up with a better theory.
I think most immigrants become either extreme winners, extreme losers/going on benefit or become extreme hard workers. The situation of immigrants leave few medium choices.
1. Start with a dumbed-down version of an opposing position: immigration causes crime.
2. Refute it using propositions that few serious opponents would dispute: immigrants are heterogeneous; socioeconomic conditions matter; migration can have benefits; crime has multiple causes.
3. Present that refutation as though it settles the substantially harder disputes people are actually having: the effects of the scale and composition of migration, integration outcomes, specific forms of crime, fiscal effects, segregation, and the effectiveness of particular policies.
Weak.
Britain provides an even more amusing historical experiment.
Between 1788 and 1868 Britain and Ireland transported about 160,000 convicted people to Australia. Many had been convicted of theft and other property offences.
https://www.nationalarchives.gov.uk/help-with-your-research/research-guides/criminal-transportation/
Britain therefore literally tried exporting offenders.
It did not export crime.
Crime remained a major social problem in nineteenth-century Britain because the conditions that generated theft, violence and disorder remained: urban poverty, insecure employment, overcrowding, extreme inequality, alcohol problems and enormous social disruption during industrialisation.
That example reveals the central error in treating crime as something residing inside particular populations.
You can move people.
Unless you change the conditions, other people can occupy the same social position.
This does not mean that poverty mechanically turns people into criminals. Most poor people never commit serious crimes. Nor does inequality excuse a robbery or shooting. Explanation and moral justification are completely different things.
But criminology asks why the probability of offending differs between environments.
Here inequality, deprivation, social stress, segregation, criminal opportunity, weak social trust and access to legitimate employment repeatedly matter.
Immigration can interact with these variables. A very rapid refugee inflow can strain housing, schools and local services if government refuses to expand capacity. Poor integration policies can create unemployment and segregation. Particular criminal networks can develop within particular migrant communities.
Those are real problems.
But notice what has happened to the argument.
We are no longer saying:
immigration → crime
We are saying:
labour-market exclusion + inequality + segregation + age structure + stress + criminal markets + institutions + policy → crime risk
Migration is one possible input into that system. It is not the system itself.
So the opponent's third point does not rescue his argument.
Yes, investigate scale.
Yes, investigate composition.
Yes, investigate integration.
Yes, investigate particular crimes.
Yes, investigate fiscal effects.
But those are empirical questions, not evidence that immigration causes crime.
And fiscal effects do not logically prove a criminological effect at all.
The scientifically defensible conclusion is therefore more modest than “inequality explains all crime”, but considerably stronger than the anti-immigration position:
Immigration is not a general or sufficient explanation of crime. Socioeconomic deprivation, inequality, segregation, social stress, criminal opportunity and institutional conditions have much broader historical and empirical explanatory power. Migration sometimes affects crime because it interacts with those conditions. Good integration policy can therefore reduce the risk, while austerity, exclusion and concentrated poverty can increase it.
That is not avoiding the difficult question.
It is answering it.
The criticism sounds rigorous, but it does not actually refute my argument.
It is true that serious research must examine the scale and composition of migration, integration, segregation and different types of crime. But simply listing those variables does not show that immigration is the cause of crime. They have to be tested empirically.
And when researchers do that, the picture becomes much more complicated than “more immigration = more crime”.
Sweden's National Council for Crime Prevention, Brå, reviewed Swedish and international research in 2025. The report was produced following a government assignment and therefore emerged in a political context in which the Tidö government had explicitly made reducing immigration a central objective. That is worth mentioning. But there is no evidence that the government ordered Brå to reach a predetermined conclusion.
More importantly, Brå did not establish such a conclusion anyway.
Brå writes that there is no simple relationship between migration and national crime trends. International research shows different results for different immigrant groups, countries, periods and crimes. Some studies find increases in particular offences; others find no significant effect, and some find lower crime associated with immigration.
Brå also stresses that the effects of migration are small compared with all the other factors affecting crime in both migrant and native populations.
Source: https://bra.se/rapporter/arkiv/2025-05-16-migration-och-brott
That is already enough to expose the logical problem in the original criticism.
Nobody disputes that the composition of a population matters. A population containing many young men will normally have more crime than an older population, regardless of ethnicity. Refugees arriving after war may initially have worse labour-market conditions than recruited engineers. Migrants placed in segregated neighbourhoods with poor schools and unemployment face different conditions from migrants entering employment immediately.
But these observations weaken rather than strengthen a simple ethnic explanation.
They tell us that conditions matter.
The same principle becomes particularly obvious when we examine native-born Swedes.
Sverigedemokraterna is a useful control case. I am not claiming that being a Sweden Democrat causes criminality. That would reproduce exactly the statistical error I am criticising.
But SD candidates have repeatedly been unusually overrepresented in Swedish investigations of criminal records.
In a 2018 review, more than one in twenty SD candidates had been prosecuted for crime during the period examined, the highest share among the parties. Another local SVT investigation in Örebro County found that roughly one in five SD candidates had criminal convictions; offences in the candidate population included aggravated assault, narcotics offences and attempted arson.
https://www.svd.se/a/7lO5o4/sd-toppar-lista-over-andel-atalade-kandidaterhttps://www.svt.se/nyheter/lokalt/orebro/var-13-e-politiker-ar-domd-for-brott
The pattern did not disappear afterwards. Acta Publica found 64 candidates from parliamentary parties with links to criminal motorcycle environments before the 2022 election. It also identified 289 candidates who had expressed Nazi or racist ideas or had similar far-right connections; 214 of those candidates represented SD.
https://actapublica.se/nyheter/makten-nasta/https://actapublica.se/nyheter/rapport-ur-mylla-brun/
In 2026, another large investigation found criminal convictions during the previous ten years among 3.3 per cent of SD candidates, compared with 1.1 per cent among Social Democratic, Green and Centre Party candidates.
https://www.etc.se/story/stor-granskning-sd-har-valets-mest-kriminella-politiker-8tycdfu6
Again, the correct conclusion is not that Sweden Democrats are inherently criminal.
The interesting conclusion is the opposite: statistical overrepresentation does not explain its own cause.
We therefore have to ask why particular groups recruit or contain more people with certain social backgrounds.
That principle must apply equally to immigrants.
The history of SD itself also makes simplistic talk about a uniquely immigrant culture of violence rather awkward. During the 2010 incident later known as the “iron-pipe scandal”, leading SD representatives Erik Almqvist, Kent Ekeroth and Christian Westling became involved in an aggressive confrontation in Stockholm. Video showed racist and sexist abuse, and Almqvist and others armed themselves with metal pipes.
https://www.svt.se/nyheter/inrikes/sd-topparna-bevapnade-sig-med-jarnror
More recently, violent far-right networks known as Aktivklubbar have grown in Sweden. These are not representative of ordinary conservative voters and should not be confused with the democratic right. They belong to the neo-Nazi and militant far-right environment. In 2025 four members of Aktivklubb Sverige received prison sentences after unprovoked attacks on people selected because of their foreign background.
https://expo.se/granskning/valdsdomda-aktivklubb-aktivister-hyllas-borde-fa-en-medalj/
There is an additional irony. A leading member of the criminal motorcycle gang Comanches attended Jimmie Åkesson's wedding in 2024. Åkesson said afterwards that he had not known about the man's gang connections. That qualification matters. But Swedish police nevertheless described the relationship as problematic because organised motorcycle gangs actively seek political influence.
https://www.svt.se/nyheter/inrikes/mc-topp-var-gast-pa-jimmie-akessons-brollop-sd-ledaren-misslyckandehttps://www.svt.se/nyheter/inrikes/polisen-om-akessons-koppling-till-mc-topp-problematiskt-de-soker-politiskt-inflytande
None of this proves anything biological or cultural about right-wing Swedes. Precisely.
It shows why we should not make that inference about immigrants either.
There is much stronger evidence for socioeconomic mechanisms.
Wilkinson and Pickett's The Spirit Level became famous for showing that more unequal affluent societies tend to experience worse social outcomes, including more violence. Their original work has attracted legitimate methodological criticism, so it should not be treated as the final word. But subsequent research has repeatedly found an association between income inequality, low social trust and homicide.
One cross-national study of 33 countries found homicide and assault to be more common where income inequality was greater and identified social trust as an important mediating mechanism.
https://academic.oup.com/eurpub/article/21/2/241/498070
A 2024 re-examination of the Spirit Level thesis using updated data and additional controls also found substantial continuing evidence connecting inequality with social harm, although individual outcomes vary.
https://journals.sagepub.com/doi/10.1177/00207152241245620
Austerity gives us even stronger causal evidence.
Researchers studying Britain's Welfare Reform Act of 2012 compared places differently exposed to welfare cuts. Areas hit harder by the cuts subsequently experienced larger increases in crime, driven especially by violent crime. The effect was strongest in neighbourhoods that were already deprived.
That is much closer to a natural experiment than simply correlating ethnicity with offending.
https://arxiv.org/abs/2012.08133
History also destroys the idea that modern immigration is necessary to produce high levels of violence.
Swedish historians estimate that homicide rates in the fifteenth and sixteenth centuries may have been around twenty times modern Swedish levels. Sweden obviously did not have today's refugee immigration, multicultural suburbs or Middle Eastern gangs.
https://www.scup.com/doi/pdf/10.1080/14043850050116264
European homicide subsequently declined enormously over several centuries. That decline began before the modern welfare state, so it would be wrong to credit the welfare state with all of it. Historians point to stronger state institutions, declining private violence and changing social norms among the causes.
But the historical evidence establishes something simpler and very important: large-scale violence existed perfectly well without modern immigration.
…
It may be true that non-European immigrants do best in higher education, but they also do "best" in crime. I wonder why. Are there two distinct groups going two different roads?
A private theory is that those who come for working do better, and those who flee from violence do worse. Those who come for work take a step to a better life, those who flee from violence may end up in desperation.
I would like the theory to be tested. Or if that has already been done I would like someone coming up with a better theory.