Abstract: | This essay reports results on optimal growth in a two‐sector model with fixed coefficients, irreversible investment and no discounting. Under normalization, the model can be represented by two real numbers, but despite its deceptive simplicity, it admits rich transition dynamics and apparent pathologies that seem to have been missed in earlier work. From a methodological point of view, and in the light of recent work of Nishimura and Yano, this essay can also be seen as a further rehabilitation of geometric methods as an engine of analysis. |