The October 2006 issue of Psychological science contains two articles about the topic of race and intelligence. William Dickens and James Flynn argue that the IQ gap in the U.S. has shrunk by 4-7 points.
However, another article by J. Philippe Rushton and Arthur Jensen in the same issue is based on a more comprehensive source of data. Their result is that the IQ gap has remained between 15 and 18 points which is 1.1 standard deviations (of the normal distribution of the general population). The constancy of this number, first measured during the First World War, is sometimes referred to as the fundamental law of sociology.
Prof. Rushton says that the origin of the differences is 80% genetic and this fact wouldn't allow the IQ gap to shrink by more than 3 points during the era - numbers that the Reference Frame is not able to check or debunk.
CapitalistImperialistPig has pointed out an interesting June 2006 article by Steven Pinker about the interactions between social phenomena, genes, and intelligence of the Jews.
The previous article about a similar issue was dedicated to the 3.6-point IQ difference between the genders.