Tuesday, August 16, 2005

Flynn Effect in Denmark & Norway - R.I.P.?   posted by Theresa @ 8/16/2005 05:00:00 PM

I read with interest (and, admittedly, a hell of a lot of ignorance) both the Teasdale & Owen and Sundet, Barlaug & Torjussen articles which purport to show an end to the Flynn Effect in Denmark and Norway respectively (see A. Beaujean's previous post for more on the topic).

Now, I am far from an expert on the ins and outs of IQ and the Flynn Effect -- and I will leave it up to others to discuss the likelihood of the Flynn Effect coming to an end in industrial countries. However, what I do want to do is to call in to question the methodology behind both of these investigations. Specifically, the validity of the researchers' data sets.

In both cases, the researchers examined data from intelligence tests given to military conscripts over several decades (Denmark and Norway have mandatory military service). In the Danish study, the data ranges from 1959 through 2004; in the Norwegian, from 1957 through 2002. Again in both cases, test scores increased from the 1950s onwards (at varying rates) through the 1980s until they reached a peak in the 1990s at which point they appear to decline. Thus, the researchers conclude an end to (or at least a slowing of) the Flynn Effect.

Each of the articles goes on to discuss various possible alternate causes for the differences between the cohorts from the various years -- they examine differences in nutrition, education, etc., etc. But in neither case do they question the obvious -- i.e. that their data sets may have changed.

Anyone who knows anything about Europe should know that the populations of both Denmark and Norway were very different in the 1950s as compared to today. Due to fairly steady immigration since at least the 1970s, neither of these nations have homogeneous populations any longer. (1) (2) Furthermore, not only are the populations of today very different from those of several decades ago, perhaps it is not coincidental that the dip in the intelligence scores in both countries happened in the 1990s -- presumably just as the children of the immigrant groups of the '70s and '80s came of age and entered military service. (3)

Of course, the next question would be, could immigration have affected these test scores? -- and by how much? I don't know for sure -- but, if I were to try to work out some answers, I would certainly consider the following:

  1. What populations are we talking about here? The largest immigrant population group in Norway is Pakistani; in Denmark, it appears to be Turkish, although there are also large groups from Sri Lanka, Iran, Iraq and Africa. (4) (5)
  2. What are the median IQ scores for these groups? Lynn & Vanhanen in "IQ and The Wealth of Nations" estimated the median IQs for Pakistan to be 81, Turkey 90, and Sri Lanka 81. (6) What happens when we factor in such median scores with the median score of 98 for both Norwegians and Danish?
  3. How many young men from immigrant backgrounds are entering the military service at any given time? Perhaps the armed forces of each country (the agencies who gave the tests in the first place) have some numbers. A very rough estimate of mine for Denmark in 2004 suggests that approx. 10% of Danish conscripts may have come from a non-Western immigrant background. (7) What sort of implication does that sort of number have for the overall test scores?

All but one of the five researchers involved in these two studies are (or at least were) based in Norway and Denmark. I find it hard to believe that they are ignorant of the changes in the populations of the countries in which they work/live. I find it even more unfathomable that they didn't take such facts into consideration when they undertook their research.


(1) "At the beginning of 1970, the immigrant population in Norway totalled 59,200 persons, which was about 1.5 per cent of the total population. By the beginning of 2004, this figure had increased to 348,900, 7.6 per cent of the Norwegian population.... In 1970, people with non-western origins accounted for 16 per cent of the immigrant population, while in 2004 the figure was 72 per cent."

Immigration and Immigrants 2004

(2) "In 1980, there were about 135,000 immigrants in Denmark out of a population of about 5 million. This figure has increased rapidly since the mid 1980s and in 1998 there were 277,000 immigrants in Denmark, corresponding to 5.2 percent of the Danish population. These figures include 1st and 2nd generation of immigrants, where the latter group has been steadily increasing from 18,000 in 1980 to 70,000 in 1998."

Employment and Wage Assimilation of Male First Generation Immigrants in Denmark

(3) "Almost half of the persons in the immigrant population [in Norway] were aged 20-44 years, whereas the corresponding figure for the total population was 35%.

Immigration and Immigrants 2004

(4) At the beginning of 2004, the immigrant population in Norway totalled 349,000 persons, almost 8% of the total population; first generation immigrants totalled 289,000 persons, or 6% of the total population. Almost 3 out of 4 persons in the immigrant population had non-Western backgrounds. Two out of 3 first generation immigrants also come from a non-Western country. The non-western immigration population makes up almost 6% of the Norwegian population. The largest groups in the immigrant population were persons with backgrounds from Pakistan, Sweden, Denmark and Vietnam; the largest groups of first generation immigrants - Swedes, Danes, Pakistanis and Iraqis.

Immigration and Immigrants 2004

(5) Employment and Wage Assimilation of Male First Generation Immigrants in Denmark

(6) http://www.gnxp.com/MT2/archives/002608.html

(7) There were an estimated 2600 18 year-old men from a non-Western background in Denmark totally in 2004. In the Teasdale & Owen article, they give 23,505 as the number of conscripts tested by the military in 2004. They suggest that up to 10% of 18 year-old men are exempted from military service annually; so, using this estimation, only 2340 of the 2600 non-Western men would actually be in the service in 2004 -- or approximately 10% of the total.

Statistics Denmark