68 Books & arts The EconomistJanuary 30th 2021
21Johnson Stress tests
How to design language tests for citizenship—and how not to“P
erfect swedishis overrated. But
comprehensible Swedish is deeply
underrated,” says Ulf Kristersson, the
leader of Sweden’s centre-right Moderate
party, which supports a language re-
quirement to become a Swedish citizen.
The left has come round, too: the Social
Democrat-led government plans to
introduce a language test. Sweden would
thereby leave the small club of European
countries that do not make passing such
a test a condition of naturalisation.
To learn the language of the country
you live in is the key to a full life there.
But many experts in language policy
oppose testing for citizenship—because
they suspect a less compassionate mo-
tive in some who propose them. “Becom-
ing a Danish citizen is something one has
to become worthy of,” said Inger Stojberg
in 2015, when she was the immigration
and integration minister in Denmark’s
centre-right government—implying that
the unworthy had been slipping through.
Her thinly camouflaged goal was not to
improve immigrants’ Danish, but to
naturalise fewer of them.
And so the Danish government,
which already had language require-
ments, tightened them significantly. To
prove they had reached the specified
level in a recent test, applicants had to
skim 16 pages of readings on the “People’s
Enlightenment”, a movement originat-
ing in the 19th century to give ordinary
Danes self-improving institutions such
as evening classes, libraries and scout-
ing. Applicants must answer questions
like: “In principle the People’s Enlight-
enment is for adults, but children can
take part in classes intended for families.
But what are the requirements for chil-
dren to take these classes?” Though not
exactly Kierkegaard, the material is well
above the level needed to get by.to find the time and energy for classes.
Some are barely literate; answering fairly
sophisticated written questions means
first learning to read and write. When
immigrants are told they must meet a
highly demanding standard, many stop
trying, say language teachers and re-
searchers. They may then remain isolat-
ed in their ethnic communities, the only
ones that will accept them.
Ricky van Oers, a Dutch scholar affili-
ated to the Western Norway University of
Applied Sciences, explains the effects of
raising the bar in her home country.
Once, knowledge of the language was
assessed in an informal conversation
with a local official. When in 2003 writ-
ten tests were introduced, successful
applications fell by half. Since 2007
classes in Dutch have been left to a mix of
state and poorly regulated private
schools. Reaching the necessary level is
estimated to cost €3,600 ($4,380) on
average, generally paid by the immigrant.
Today, new arrivals have three years to
reach that standard—one admittedly
more modest than in Denmark—or face
being fined up to €1,250 and being barred
from permanent residency and citi-
zenship until the test is passed.
Countries that introduce language
tests for citizenship should make sure
teaching is of a good quality (the stu-
dents themselves are in a poor position
to spot a dodgy operator). They should
learn from Germany, which subsidises
lessons so generously that they are prac-
tically free. Help people attend by ensur-
ing child care if they need it. Provide
incentives such as assistance with work
placements. Presented with a feasible
goal that can be reached with better
skills, newcomers will work harder than
when ordered to scale a distant peak that
they can scarcely see.The trend in the West is clearly towards
such tests. America and Britain typically
require English for citizenship—in 2019
Donald Trump proposed adding require-
ments for certain visas as well. But the
problem seems especially acute from a
small-country perspective. Many Euro-
pean countries are linguistic communi-
ties. Europe is a crowded continent where
neighbours often distinguish themselves
primarily by how they speak. Centuries of
nation-building from the top down
strengthened the association of one lan-
guage with one people in one state, at least
in the ideal case. English already threatens
the role of small languages. If Denmark,
say, does not require even Danish citizens
to speak Danish, what is the language for?
Even the keenest advocates of immigra-
tion agree that speaking the language
helps migrants integrate. But they argue
that the focus should be on helping them
do so, not on overly ambitious targets
many can never achieve. Many of Europe’s
migrants today arrive with psychological
problems born of fleeing war and catastro-
phe. Others work so hard that they struggleideas. Though they generally pay their way,
immigrants are often perceived as a bur-
den, he observes. This is because public
services respond too slowly to changes in
population, so schools and doctors’ wait-
ing rooms become crowded. This suggests
that governments should learn from the
private sector, which adapts far more
quickly. “Nobody blames migrants for
shortages at local supermarkets—because
there aren’t normally any,” he notes.
Facts matter, but so do words, he argues.
People who oppose immigration in the ab-
stract often change their tune when askedabout specific types of immigrant, such as
doctors, students or foreigners who marry
locals. As for those who complain about a
“brain drain” from poor countries—this is
like labelling women’s entry to the labour
force the “family abandonment rate”.
“Wretched Refuse?” is a denser book,
full of charts and regression analysis. It is
also highly original, and takes a chainsaw
to the most intellectually respectable case
against immigration. No serious econo-
mist denies that when people move from
poor countries to rich ones, they become
more productive and their wages soar. Itseems likely, therefore, that more migra-
tion would make the world much richer.
However, some scholars think that too
large an influx from, say, Congo to Canada
would make Canada more like Congo—ie,
the immigrants would import Congolese
habits, and gradually make Canadian insti-
tutions more like the corrupt and lawless
ones that keep Congo poor. Yet remarkably
little effort has been made to test this hy-
pothesis. Messrs Nowrasteh and Powell do
so as rigorously as they can.
Disentangling cause and effect is tricky.
Simply pointing out that countries with