He combined neo-Ricardian and Marxist ideas with Keynesian aggregate demand and he thought marginalist neoclassical economics was incoherent, for reasons related to the Cambridge capital debates. Robert Vienneau offers one summary of his work. Matías Vernengo writes:
As early as 1961, while spending an academic year at MIT, he suggested during a presentation by Paul Samuelson that his results depended on the assumption that all sectors use the same capital-labor ratio. The final results of his critique were presented in Garegnani’s paper “Heterogeneous Capital, the Production Function and the Theory of Distribution.” His paper shows conclusively that the marginalist theory of value and distribution based on an aggregate production function is untenable. This of course builds on Sraffa’s work in the Production of Commodities (PC). By 1966, in the famous Quarterly Journal of Economics (QJE) Symposium, Samuelson had admitted that the neoclassical parable was not defensible.
Here are some of his articles on scholar.google.com. He always struck me as a very intelligent writer and capable of a good bracing critique, even though I don’t think his proposed alternatives have gone anywhere. Every profession needs smart and articulate dissenters and I am glad that we have had Garegnani.
For the pointer I thank Juan Carlos Esguerra.