This thesis tests two explanations of the variation in restriction of managerial discretion imposed by preference share contracting, these being the nature of the assets in which the firm invests, and taxation incentives. It is hypothesised that firms with higher proportions of assets-In-place are more likely to issue preference shares with more debt-like attributes, more restrictive covenants and more covenants based on accounting information. This tested at a cross-section in time, and according to the change in assets-in-place, as a proportion of value, across time. There is limited evidence in support of this hypothesis. It is further hypothesised that firms were more likely to issue redeemable preference shares between 1972 and 1986, during which time it is hypothesised that preference shares were a tax minimising alternative to debt for identified tax clienteles. Evidence strongly supports this hypothesis. Survey evidence as to the form and restrictiveness of covenants is presented, suggesting that as the number of debt-like attributes increase, there is a corresponding increase in restrictiveness of covenants, consistent with Australian evidence on debt contracts.