A natural generalization of least squares is proposed to estimate parameters in simultaneous linear equations. Full-information maximum likelihood is shown to be identical with this generalization. The extent to which other estimators deviate from the generalization is discussed. A paradox of Strotz is resolved, and application of canonical correlation theory to structural equations is indicated. 1. SUMMARY THIS PAPER compares various estimators of the parameters of linear simultaneous

This paper gives a non-linear growth model, which explains the development of an economy through stages somewhat similar to the Rostovian stages. Non-linearity is introduced by including the inaugmentable factor of land or natural resources in the production function along with labor and capital, and by recognising that net saving is not a linear homogeneous function of income alone, but might be affected by the distribution of income and the interest rate and tends to be negative when per capita income is very low. Furthermore, population growth is assumed to follow a NeoMalthusian pattern. The effects of non-neutral as well as neutral technical progress are discussed in this paper.