CAR CRASH

If the category changes halfway through, the argument has changed its subject.

The OECD separates work, family, humanitarian and free-movement migration. CBS distinguishes nationality from birthplace-based origin. A table about foreign citizens cannot automatically describe Dutch-born children of migrants. A claim about asylum cannot borrow a total that covers every reason for moving and keep the same meaning. Name the country, group, period and outcome. Is this arrivals during a year or residents already living here? A first job or a living wage? Recorded suspicion or a conviction? Only then ask whether the study describes a difference or actually tests an explanation for it.

Watch what happens when an origin category quietly becomes a description of belief. Someone is counted because of a birthplace, then discussed as if the table had measured their religion, loyalties or values. The number did not acquire those extra variables on its way into a caption. If that is the question we want answered, we need evidence that actually asks it.

This is an argument using selected Dutch sources and OECD comparisons, not a systematic review or a calculation of migration’s total effects. Those limits do not make every interpretation equally good. They tell us which conclusions these particular records can carry and which require additional evidence.

Take recorded differences seriously without giving them a power they do not have.

CBS reports differences in registered suspicion by origin. Adjusting for age, education and income makes the gaps smaller, not universally absent. I will not pretend the differences are imaginary. I also will not pretend a suspect register is a court judgment or that adjustment establishes every cause.

Use the hypothetical group in the pictogram. At 2%, twenty of one thousand people are registered as suspects; 980 are not. Compared with 1%, that is twice the rate. Both descriptions matter. Neither proves that the registered twenty are guilty, let alone the remaining people.

Ask what changes over time before treating the first snapshot as the whole person.

CBS’s 2026 release finds newer refugee-permit cohorts entering work faster, while the groups and circumstances also differ. A difficult first period therefore cannot settle a lifetime’s contribution. Improvement does not prove which policy caused it, or mean everybody has reached stable work and financial independence.

Decide what success means. Time to a first job, hours worked, a wage someone can live on and use of a qualification are different outcomes. A tiny temporary contract can be progress without solving the practical problem. Count the result you actually promised to improve.

Consider a hypothetical newcomer who has found work but cannot yet use their qualification. One headline can present employment as the finish line; another can present the unused qualification as proof of failure. Neither describes the whole situation. I would ask what changed, what obstacle remains and which next step could improve the result. That leaves us with a person and a practical problem we can work on.

Compare language learning, childcare, qualification recognition and access to employment against costs and outcomes. My preference is to make participation possible, expect an honest effort and change measures that fail. That is a policy proposal to test, not a result already proved by a rising line.

An inconvenient cost deserves scrutiny. A dramatic headline does too.

The OECD’s 2021 fiscal study covers 2006–2018. Its results depend on population composition and accounting choices. Read it as historical evidence, not today’s invoice for new arrivals. Ask who is included, what years are followed and what the comparison without migration assumes.

A one-year balance and a lifetime estimate answer different questions. Assigning shared public spending and assuming future employment also changes the calculation. Show those choices beside the total. Methodological scrutiny should explain the result, not become a convenient excuse for discarding a number you dislike.

An accounting choice is worth making concrete. If a study allocates a share of an existing public expense to each resident, that allocation is not automatically the same as the extra expense caused by one arrival. The study may explain why its choice is useful. Read that explanation before carrying the total into a different question. Comparisons become more honest when their assumptions travel with them.

Public money matters because bad spending leaves less for someone who needs it. But I would not make an elderly relative, a disabled neighbour or myself during a bad year earn fair treatment through a positive tax balance. A fiscal estimate cannot finish the moral argument about a person.

A shortage is real even when collective blame is the wrong explanation.

A 2023 Dutch government letter recognises migration’s housing pressure. OECD comparisons also show migrants disproportionately experiencing overcrowding. Telling a waiting resident to stop worrying is insulting. Presenting a newcomer in poor conditions as the person who stole everyone’s house is insulting too. Both problems deserve a response.

Open the three policy questions: demand, supply and allocation. How many households need which homes, and when? What can be built or made available? Who receives a particular vacancy under which rule? Changing that last rule is not the same thing as creating another home.

If we want participation, we need to examine the barriers as well as the effort.

A Dutch correspondence experiment found unequal employer responses to equally qualified applicants with different ethnic signals. Deliberately changing what employers see tests a recruitment mechanism more directly than simply observing two populations with different employment rates. That finding deserves attention without being inflated into an explanation of everything. An application-stage experiment does not describe every employer, occupation or career. It cannot assign the cause of every employment gap. It does justify scrutinising recruitment rather than assuming each unequal result reflects effort or ability. Proposed fixes should face the same demand for an actual evaluation.

Think of the applicant at the other end of the experiment. The qualifications are on the page; an ethnic signal changes the response. That is a concrete mechanism to investigate. It also gives an institution something it can examine and change. A general declaration that everyone should try harder leaves that mechanism untouched, just as assuming every rejection proves discrimination would outrun the evidence.

Fair treatment still leaves room for sharp criticism of misogyny, antisemitism, homophobia and coercion, whoever expresses them. Name the belief and harm. Leave room for someone to reject a custom, leave a religion or disagree with a neighbour. Appearance cannot supply the convictions you want to challenge.

An honest account keeps the finding and its limit in the same sentence.

The crime figures describe recorded suspicion and differences that adjustment reduces without universally removing. The work figures describe changing cohort outcomes without isolating one cause. Neither is an all-purpose judgment about migrant character. Each gives a more specific question for prevention, participation and further research.

The fiscal study examines a historical period under accounting assumptions. The housing evidence identifies pressures and poor conditions that planning must address. The recruitment experiment tests a particular barrier. Treating these as one single score would hide the distinct choices a useful policy has to make.

This record does not provide one correct migration quota or a universal explanation of crime. It does undermine the shortcut from a real problem to inherited guilt. We can recognise pressure, investigate behaviour and test policy without pretending an origin label makes every member the same person.

Refusing inherited guilt does not require pretending risks and limits disappear.

A reasonable objection says policy cannot examine every future case individually before making any general rule. Agreed. Group information can help identify pressures and guide prevention. The question is which measure follows, what mechanism makes it work and how its consequences and obligations will be assessed.

Wanting less migration is not automatically racism; welcoming more is not automatically wisdom. Each proposal owes an account of capacity, costs and protections. Neither position answers an accusation against a neighbour who did nothing. Risk information and personal guilt are doing different jobs in the argument.

A useful proposal should survive a question from the person who pays its cost. What happens to someone denied entry, a resident still waiting for housing, an employer who recruits workers or the municipality expected to provide services? Their interests can conflict. Stating the conflict openly gives us a policy to debate. Hiding it behind either generosity or toughness gives us an identity to applaud.

I also reject the comfort of explaining everything through one favourite cause. The sources cannot make every gap discrimination, every difficulty bad character or every shortage migration alone. State the problem narrowly enough that a result could show your policy failed. That is how disagreement becomes useful.

Demand practical policy and individual responsibility with the same seriousness.

The attacker in the hypothetical video still needs to be stopped and investigated. The victim still deserves protection. Nothing in this essay excuses the act. What changed is the claim about everyone outside the frame: a shared background did not connect those people to this offence.

My standard is practical: reduce harm, make participation possible, state capacity honestly and check whether promised results arrive. Argue about admission levels and the distribution of costs. Criticise institutions, ideologies and conduct directly. Make the proposal precise enough that its success or failure could become visible.

There are two people I want the argument to remember: the victim whose harm must not be explained away, and the stranger who has been dragged into the story. We owe the first a serious response to what happened. We owe the second an accusation supported by something they actually did. Holding both thoughts together is harder than choosing a camp, and considerably more useful.

State your conclusion, the source that supports it and the strongest challenge. Then ask what action follows. If the answer is only to distrust anyone with a certain face, return to the missing step. An innocent person should not have to disprove somebody else’s biography.