H-48

Appendix D Note on Proposed Treatment op Housing-costs in the Index 1. It is the general practice in other countries, and has hitherto been in this also, to confine the study of housing-costs to rented houses. This is easily understood. Rents are fairly easy to collect white the collection of statistics as to the outgoings on owner-occupied houses presents considerable difficulty. It may, too, with some show of justice be argued that the movement in rents should fairly represent the movement in landlords' costs; and this again should approximate to the movement in owner-occupiers' costs. 2. Rent control has, however, altered the picture, in that average rents of rented houses may no longer be typical of housing-costs in general; while, again, State houses rented as a matter of State policy at low rentals form an increasing proportion of rented houses. Attempts have therefore been made to devise a measuring-rod for housing-costs. Fresh ground has had to be broken, but it is probably correct to state that the need of such a new technique is greater in New Zealand than elsewhere on account of the comparatively large proportion of houses which are owner-occupied. he 1945 census revealed that only 35-38 per cent, of the private European dwellings in New Zealand are rented. Of the remainder, 7-73 per cent, are rent-free (mostly with job), 31-36 per cent, are owneroccupied subject to mortgage, and 25-38 per cent, owner-occupied without mortgage (a small percentage were not specified). a & \ s 3. As with food and other items in the regimen, the weight for housing-costs was determined on a national expenditure basis with a primary division into three components : (1) rents of houses, (2) rents of flats, and (3) expenses of owner-occupiers : - (1) Rents at present being collected from twenty-five towns provided a basis of assessing total rent payments. The towns were classified into (a) four main centres, (&) other urban areas, and (c) other towns. In each of these classes an average rent of four, five, and sixroomed houses was calculated. (The weights for combining these three sizes of houses were m every case the respective numbers of all four, five, and six-roomed houses at the 1945 census.) A small adjustment provided a conversion from the average of four, five, and sixroomed houses to the average of all houses. This average for each class was then multiplied by the number of rented houses (excluding flats) in the class—i.e., (a), (&), and (c)—in turn. In the case of class (c), however (other towns), the number of houses used included a so houses in rural areas, the index being designed to measure movements in urban costs ° nly ' F ° r the ab ° Ve purposes ' houses rent-free with job were reckoned as rented houses. ill ™! atS W6re Similarly treated > the sizes used being in this case two, three, and four rooms. (3) lhe gross rental value of owner-occupied houses had already been investigated as part of a national income survey, and data revealed in that computation were adapted to this purpose. (The computation was based on (a) estimated numbers of owner-occupied houses in 1946 (from 1945 census), excluding, of course, rent-free houses already taken in, and (b) rental values derived from 1936 census, increased by 20 per cent, to allow for the fact that it is generally the better class of house that is owner-occupied, and also adjusted by the movement in the present rent index from 1936 to 1946.) While the proportion of total flats to total rented dwellings was shown by the 1945 census to be approximately 1 : 4 in twenty-one towns combined, the proportions vary considerably as between various towns, so that this factor has to be taken into account. To obtain a properly weighted combined rents and flats index, house and flat rents will be combined in the following proportionsWellington : Rented houses, 2 ; flats, 1. Auckland, Christchurch, Dunedin, Hamilton New Plymouth, Palmerston North, Nelson, and Timaru : Rented houses, 3 ; flats, 1. Other towns : Rented houses, 5 ; flats, 1. The index for each town for rented houses and flats will then be combined with the index of costs of owner-occupiers to obtain the housing index. This will be effected by using the following proportions : Rents (including flats), 40 per cent.; costs of owner-occupiers 60 per cent. The New Zealand index for housing will be compiled by combining the indices for each of the twenty-one towns by the population weights shown in Appendix C. 4 Passing on now from weighting to pricing, groups (1) and (2) present no difficulty. The collection of rents of houses and flats will proceed as in the past, only with the revised geographical coverage (twenty-one towns, instead of twenty-five). The assessment of the changes in expenses 01 owner-occupiers, however, requires special care. The constituent elements of these costs were

35