We explore the basis for the data on imputed rental of owner-occupied dwellings for the period 1948-1985. In the current UK national accounts dataset, these estimates are based the assumption that the revisions which took place in 2016 UK National Accounts and 2017 UK National Accounts were assumed to build up gradually from 1948 onwards, so that revisions to the early data were minimal while those to the more recent data were substantial. We show that the current data imply very low and unlikely yields on housing in the 1950s and 1960s. Applying the 1985 pro rata adjustment to imputed rental gives a stable yield of around ten per cent which leads to a very stretched ratio of imputed rental to actual paid rental per dwelling in the 1950s. Accordingly, it is not possible to come to a definite conclusion other than that more investigative work is needed. We suggest that the ONS convenes a panel of experts to undertake further work to establish the best way forward and establish a more plausible dataset of better quality.
