An econometric analysis of farmland values in Western Canada

Abstract

Interest for this research was initiated by the threefold increase witnessed in rural land values since the early 1950's. Society in general and farmers, mortgage lenders, tax assessors and public policy makers in particular are concerned with these price increases because of their economic, social and political implications. The major objective of this study was to identify the more important factors influencing farmland prices, develop a methodology to measure their individual effect on market value and, if possible, project future land prices for alternative values of these determining variables. Before hypotheses concerning sources of price variation were generated or functional relationships specified, familiarization with the nature and structural characteristics of the farmland market itself was accomplished by personal observation of the market in operation, discussions with participants (buyers and sellers) and a fairly exhaustive review of major studies on land values in both the United States and Canada. Two econometric models of the farmland market were subsequently formulated -- a cross sectional or static model- and a time series or dynamic model. Four basic hypotheses were tested -- the market value of farmland is positively related, to (a) soil productivity, (b) expected net farm income from land, (c) long term farm credit availability and (d) technological advance in the agricultural industry. Land value data for the empirical analysis were obtained from two independent sources -- (a) Manitoba Crop District 3 and 10 sales data (1958 - 1970) and (b) D.B.S. estimates of the average annual "value per acre of occupied land" for each of Manitoba, Saskatchewan and Alberta. (1928 - 1969) Both models were fitted. to the economic data by ordinary least-squares linear regression...