[1972]; Westergaard and Resler [1975]; Reid [1977]; Abrams [1978]). This is partly due to an unfortunate disinclination to undertake large-scale empirical research; partly due to an exaggerated suspicion of social measurement; and partly due to an excessive distrust of officially-produced numerical data. We do not read or honour Ogburn (who more or less invented social indicators), Steinmetz (who coined the word ‘sociography’) or Morgenstern (economist and author of one of the finest critiques of error in official data [1963]) because there is little interest in the sorts of data which they sought to exploit or the uses which they tried to make of them. This paper seeks to redress the imbalance by putting forward some arguments in favour of the use of official statistics.