HRS International

Is Statistics Norway violating research ethics?

The newsweekly Ny Tid has found educational statistics from Statistics Norway that date back to 2008 and that they believe indicate that young men with immigrant background are ”at the top educationally.” Statistics Norway’s Lars Østby is backing them up on this. Ny Tid wonders why no other media have reported this ”news” before. But these statistics – and especially Statistics Norway’s presentation of them – raise more questions than the editors of Ny Tid seem to recognize.

Rita Karlsen, HRS

In fact, I’m not so surprised that the media in general have been rather reserved in presenting Statistics Norway’s research. Many people have picked up on the fact that Statistics Norway’s presentations can contain a political bias that requires one to look beyond the numbers oneself in order to guarantee a balanced account. And this takes time.

When it comes to statistics about immigration and integration, Statistics Norway has repeatedly distinguished itself with analyses and conclusions which maintain that ”integration is going well” and ”things will take care of themselves in time.” One of Statistics Norway’s leading practitioners of this sort of thing over the years has been Lars Østby. Now it can seem as if Østby’s attitudes have undergone a kind of inflation; it seems as if Statistics Norway produces studies that will and must conclude with Østby’s hypotheses.


HRS is among those groups and individuals that have been frowned upon by Statistics Norway when we have trespassed on ”their turf.” It began in the early 2000s when we studied Statistics Norway’s available figures and analyses about marriage. We noticed quickly that there was a lot there that was not being said. We thus placed a special order for statistics from Statistics Norway that could give a more detailed picture of the people that members of various immigrant groups in Norway were marrying. The figures revealed that most members of the major immigrant groups were bringing spouses from abroad. For the 12 selected population groups, during the period between 1996 and 2001, first-generation women and men (those whom Statistics Norway now calls immigrants) entered into a total of 8,478 marriages. By far the majority of the Norwegian-based individuals in these marriages wed partners from their homelands – partners who were either brought over from their homelands to marry or who were already resident in Norway. Few of these Norwegian-based individuals married partners of ethnic Norwegian origin, and almost none wed partners with foreign backgrounds that differed from their own. As for the marriage pattern among the secong generation (those whom Statistics Norway now labels Norwegian-born with immigrant parents), we noted that the non-Western immigrant population in Norway is relatively young compared to the rest of the population. This means that not very many members of the second generation had yet married. We nonetheless thought that it was important to look at those who had already married to see if they showed any marriage pattern. During the period between 1996 and 2001, there were a total of 808 marriages by second-generation women and men belonging to eight of the 12 selected immigrant groups. Second-generation Pakistanis made up the decidedly largest group that had married; they made up about 68 percent (551 marriages in all) of all the marriages entered into by those with backgrounds in the countries in question. And the tendency among the marriages that had been entered into seemed, however, to be reasonably clear – members of the second generation were mainly marrying partners brought over from their parents’ homelands (three out of four), and pretty much all of them married spouses who had the same national background as themselves and almost none married ethnic Norwegians. (See more on the statistics on our old website.)

Statistics Norway didn’t like this. For a ”typical” Statistics Norway presentation groups the figures together in such a way that they don’t reveal anything, or, to put it more precisely: they tend in the direction that Statistics Norway ”wants”. When it came to the marriage statistics that HRS presented, we were strongly criticized by Statistics Norway, which argued that the numbers were too low to be able to discern any pattern in them. In other words: all too few members of the second generation had married to make it possible to formulate any statistics and draw conclusions from the figures.

But the reason for this argument was likely that the figures did not support Statistics Norway’s own position. For when it comes to educational statistics, Statistics Norway falls into the same trap that they believed HRS fell into – but this time they maintain that the numbers prove that things are going well. And Ny Tid is buying it uncritically.

Under the heading “At the top educationally,” Ny Tid maintains in its introduction that young men with immigrant backgrounds are more likely to study than boys in the majority group. ”But this,” we then read, “has not come out previously.” And then they employ Statistics Norway as a witness to this truth with the following statement: “It is not strange that the debate is skewed, believes a Statistics Norway researcher” – who, not expected, is Lars Østby. As Ny Tid writes:

It has previously emerged that minority girls are at the top educationally, as Ny Tid demonstrated in a major report in the autumn of 2006. But today Ny Tid can present figures from Statistics Norway (SSB) which show that most minority boys, too, do like Heydari [a boy discussed at the beginning of the article]: they choose the universities over the B-gangs [immigrant youth gangs] or the “moral police” [groups of young immigrant men who police their communities for violations of Muslim law].

Among Norwegian-born men (aged 19-24) with immigrant parents, fully 31 percent pursue a higher education. This is markedly higher than the 23 percent of majority-group men in the same age group who go to college. Thus young men with immigrant background are bringing up the educational average among the Norwegian male population.

”Harmful myths”

It is now less surprising is that the same applies to minority women: 41 percent of young Norwegian-born women with immigrant backgrounds pursue a higher education, as opposed to 36 percent of majority-group girls.

“The myth of unsuccessful integration is taken as a given in TV debates and elsewhere, and one takes the discussion from there. This myth is not rooted in today’s reality, and it is harmful because it can prevent us from making the right moves,” says Statistics Norway researcher Lars Østby.

When earlier this year the spotlight was place on the problems concerning the moral police in the Oslo neighborhood of Grønland, it gave new life to the debate about the degree to which so-called integration is successful. People have strong opinions about the topic. On the newspapers’ opinion pages and on the Internet, immigration, racism, and integration are among the most frequently debated topics.

Østby thinks it is problematic that the public debate is most often founded on the assumption that integration is unsuccessful. The figures that are being published here in Ny Tid for the first time in any newspaper have been available since the autumn of 2008. Nonetheless, Østby has previously only had the opportunity to mention them in one radio interview, and the point that minority boys are at the top educationally was not picked up.

“This is a difference that is visible to the naked eye. I am rather surprised that more Norwegian newspapers have not picked up on the news,” Østby tells Ny Tid.

So let’s look, perhaps also to Østby’s “surprise,” a little more closely at the figures – and “the myths”:

According to Statistics Norway, there were 11,700 immigrants (first-generation) and 5,400 Norwegian-born people with immigrant parents (second-generation) in upper secondary schools (videregående skole) in Norway in the autumn of 2007.

The figures (for 2007) reveal that after completing primary and secondary school (grunnskole), 96 percent of all pupils went directly to higher secondary school in the next school year. The figures also show that 90 percent of all of the country’s 16-18 year olds, were in upper secondary school in the autumn of 2007. For immigrants the figure was 68 percent, and for Norwegian-born individuals with immigrant parents it was 89 percent. I will note that of the three groups (all of the country’s 16-18 year olds, immigrants, and Norwegian-born individuals with immigrant parents) not all of the groups are mutually exclusive. For if we speak of all of the country’s 16-18 year olds, surely immigrants and Norwegian-born individuals with immigrant parents are included? What does this say about the figure of 90 percent as a basis for comparison? Yes, it will pull the ”all-percent” down, especially on account of the 68 percent figure for immigrants.

We may then look at some of the real numbers: pupils with backgrounds from Iraq (1,321), Somalia (905), Russia (761), and Afghanistan (718) are the largest groups among immigrants in upper secondary school. Among Norwegian-born individuals with immigrant parents, most had Pakistani (1,346), Vietnamese (696), and Turkish (460) parents. In other words: these figures do not differ markedly from the figures we worked with in the marriage statistics – apart from the fact that it is now acceptable, apparently, to generalize and draw conclusions.

It is also interesting that Statistics Norway points out that the percentage of those who complete their education is lower, and the dropout rate is higher for immigrant pupils than it is for the entire group of pupils in upper secondary schools. The percentage that drops out is thus considerably lower for Norwegian-born individuals with immigrant parents than among immigrants. We can then read that the percentage of those who complete their education is higher among girls than among boys. Especially among immigrant boys in vocational college, there were many who quit along the way, and 55 percent of the class of 2001 dropped out in a five-year period, compared to 34 percent of all boys and 39 percent of Norwegian-born boys with immigrant parents. Again, we can see that the basis for comparison is ”all boys.”

Unfortunately, when it comes to higher education, which is the subject Ny Tid is examining, at, we can observe the same pattern from Statistics Norway as seen above: The figures from Statistics Norway show that in the autumn of 2007, 18 percent of all immigrants and 35 percent of Norwegian-born individuals with immigrant parents aged 19-24 were in higher education. By comparison, the percentage in the entire population was 30 percent for the same age group. But it is impossible, on the basis of Statistics Norway’s figures, to figure out how this last figure divides up between immigrants and Norwegian-born individuals with immigrant parents, on the one hand, and ”the remaining mass of students” on the other.

If we look at the figures for Statistics Norway’s groups ”total population,” ”immigrants,” and ”Norwegian-bon with immigrant parents” and the sexual distribution from 1997 to 2007 in higher education, we find the following:

It is from this table that Ny Tid has drawn its figures for men; compare the figures for men in the last line for 2007 in the above table, in which it is written that ”Among Norwegian-born men (aged 19-24) with immigrant parents, a full 31 percent pursue higher education. This is markedly higher than the 23 percent who study among the majority-group men in the same age group. Thus young men with immigrant backgrounds are raising the educational average among the Norwegian male population” (my emphasis).

As we see, Ny Tid has no evidence for this conclusion, since the basis for comparison supplied by Statistics Norway is ”the total population,” which includes both immigrants and Norwegian-born individuals with immigrant parents. There are, moreover, major variations when it comes to the question of who is studying. As of 1 October 2005, in all 6,016 immigrants and 2,234 Norwegian-born individuals with immigrant parents were in higher education. But this includes all national groups, as seen in this table:

Students with backgrounds from Bosnia-Herzegovina were the largest individual group among immigrants aged 19 to 24. There were also many students with backgrounds from Russia, Iran, and China. Among Norwegian-born individuals of the same age, students with parents from Pakistan, Pakistan, Vietnam, and India made up the largest groups. The percentage of Norwegian-born individuals in higher education who have immigrant parents is thus larger than the percentage of immigrants who are in higher education, but the average conceals major differences. According to Statistics Norway, measured as a percentage of the groups of individuals aged 19-24 in 2005, Norwegian-born individuals with Indian, Polish, and Sri Lankan parents clearly had the highest levels of participation in higher education in 2007, those figures being 58, 53, and 52 percent respectively. Among the largest groups, the participation rates among Norwegian-born individuals with Pakistani and Vietnamese parents were 32 and 47 percent respectively.

There are also considerable gender differences, and in all of the groups it is mostly women who pursue higher education. Among Norwegian-born women with Indian and Sri Lankan backgrounds, the percentage who pursue higher education was 64 percent in both groups, while the percentages of Norwegian-born men with these backgrounds who pursue higher education were 52 and 43 percent respectively. Among Norwegian-born people with Vietnamese parents, 50 percent of the females and 44 percent of the men were in higher education. Among Norwegian-born women with Pakistani parents, the percentage pursuing higher education was similar to the percentage for women in the entire population, 37 percent. For Norwegian-born men with Pakistani parents, the percentage pursuing higher education was 3 percentage points higher than the percentage for all men. Among young people with Turkish parents, the percentage who pursued higher education was considerably lower, 24 percent for women and 16 percent for men.

As we can see, it is anything but simple to learn anything from Statistics Norway’s presentation of statistics. If Statistics Norway were willing to produce pure, fact-based statistics, they would in all probability have compared immigrants and Norwegian-born individuals with immigrant parents with the ”majority population,” just as Ny Tid thinks Statistics Norway has done. NRK seems to have done the same thing in its reporting on the same topic – also with the help of Statistics Norway, where advisor Kristin Henriksen maintains that ”There are not so many people who know that young individuals with Pakistani backgrounds, especially boys, have been streaming to the universities.”

Statistics Norway’s representation of the facts on this subject comes close to amounting to research deceit. In any case, one may ask whether Statistics Norway’s account of these matters is ethically defensible. For it is precisely in this way that Statistics Norway misleads us time after time – and without any political action being taken. Quite the contrary: the government dares to use Statistics Norway’s figures uncritically, as Prime Minister Jens Stoltenberg did in his speech to the Oslo Labor Party committee on 4 May. ”One of the most gratifying aspects of the educational statistics is that those who are born in Norway to immigrant parents are more likely to pursue higher education than the rest of the population, and that those who are best represented in higher education are the women.”

One may really ask who it is that is creating myths.

Translated from the Norwegian by Bruce Bawer