Using sex surveys to calculate paternal discrepancy

Since the 1950s a growing literature has addressed the extent of paternal discrepancy (PD) in human populations. This literature overwhelmingly rests its claims on medical testing – initially serological testing, and from the late 1980s DNA testing. The problem with this evidence is that it is based on samples where paternity confidence is sometimes low, sometimes high, but never representative. Since the early 1990s there is a growing body of evidence on sexual behaviour which provides a new source of evidence regarding the extent of PD. This paper uses the National Survey of Sexual Attitudes and Lifestyles (Natsal) in Great Britain to calculate PD among births in 1990 and 2000. It describes procedures adopted and the assumptions made to arrive at a PD rate in the UK of 1.4-1.7% in 1990, and 1.9-2.3% in 2000 (taking the increasing use of paternity testing into account). The paper concludes that the extent of PD is much less than commonly understood, but that it is becoming more widespread notwithstanding paternity testing.