On Dharm Narain's Regression Analysis N Krishnaji This note contributes to the discussion on Dharm Narain's unfinished work on the trends in rural poverty which has remained somewhat inconclusive. It examines the statistical artifacts underlying Dharm Narain's regression fits and argues that the empirical relationships are to a large extent implicit in the manner in which poverty estimates are derived, and caused by the highly correlated trends in the different price indices used in defining the variables.