A ustralia and Canada have distinctively different macroeconomic histories. Australia's GDP per caput was well above those of Britain and the United States in 1870, and more than twice the Canadian level. By the 1980s the figure for Canada almost matched the United States, and was substantially above that of Australia and Britain. Why Canada and Australia, both regions of recent European settlement with British colonial origins, should have widely different macroeconomic records raises a number of puzzles, which we explore by utilizing new developments in growth theory and time series analysis. Our primary purpose is to investigate the extent to which Australia's links with Britain, and Canada's with the United States, shaped the comparative economic development of the two dominions. Convergence of incomes between economies arises naturally within Solow's growth model as leaders suffer diminishing returns to investment.2 However, the historical records of Australia and Canada, in comparison with those of Britain and the United States, point to a greater complexity in patterns of economic development than predicted by simple growth models, and diversity appears even greater outside the English-speaking world.3 Abramovitz argues that income convergence depends on social capability, and his perspective may shed light on the performance of the two dominions.' The possibility that income convergence may be conditional on social capability receives formal treatment in Barro's growth model, which augments the standard Solow approach to incorporate, for example, differences in education.5 Here convergence may be long drawn out, as in the case of income levels between individual US states. The issue of social capability centres on the cultural and institutional milieu underpinning an economy's technological capability. Within the augmented