After years on the increase, average intelligence test performance could be in decline. That's according to Thomas Teasdale and David Owen who took advantage of the Danish tradition of testing the intelligence of all 18-year-old men being considered for conscription into military service.
Consistent with the observed world-wide increase in average intelligence - the Flynn Effect - the 25,000 young men assessed for military service in Denmark in 1999 performed significantly better, by about 2 IQ points, than the 33,000 tested in 1988. However, the 23,000 men tested in 2003/2004 performed significantly worse than the 1998 group, at a level almost equivalent to the 1988 cohort. This apparent decline in average intelligence matches a similar observation made in Norway among their conscripts.
So what's causing this reversal in braininess? Teasdale and Owen rule out any effect of diet - after all, there's been no change in average height, which would be expected to suffer if diet quality had deteriorated.
Prior research found that test performance was higher among those men with a negative attitude towards military service. On that basis, Teasdale and Owen also reject the suggestion that the decline could be due to malingering - that is, deliberate poor performance on the intelligence test to avoid military service.
Instead, the researchers surmise that the performance decline is due to "some qualitative change in the emphasis on abstract reasoning and problem-solving within the Danish educational system or a decreased emphasis on speed". They also cite a rising proportion of immigrants in the young population as another possible contributing factor.
Finally, Teasdale and Owen noted that with average intelligence test scores beginning to rise in developing countries, the decline observed here, if representative of a larger pattern, could mark the beginning of the end for any observed differences in average IQ test scores between nations.
TEASDALE, T., OWEN, D. (2008). Secular declines in cognitive test scores: A reversal of the Flynn Effect. Intelligence, 36(2), 121-126. DOI: 10.1016/j.intell.2007.01.007