In Anthony Thirlwall’s essay Nicholas Kaldor: A Biography, 1908–1986, (first published in 1987 and republished in 2015, [note: he has a full biography too]), there’s a para which has both the history of Kaldor’s thoughts on circular and cumulative causation and a short explanation:

As Kaldor grew older (and perhaps wiser?), he lost interest in theoretical growth models and turned his attention instead to the applied economics of growth. Two things particularly interested him: first, the search for empirical regularities associated with ‘interregional’ (country) growth rate differences, and secondly, the limits to growth in a closed economy (including the world economy). The distinctive feature of all his writing in this field was his insistence on the importance of taking a sectoral approach, distinguishing particularly between increasing returns activities on the one hand, largely a characteristic of manufacturing, and diminishing returns activities on the other (namely agriculture and many service activities). Kaldor’s name is associated with three growth ‘laws’ which have become the subject of extensive debate.⁷³ The first ‘law’ is that manufacturing industry is the engine of growth. The second ‘law’ is that manufacturing growth induces productivity growth in manufacturing through static and dynamic returns to scale (also known as Verdoorn’s Law). The third ‘law’ states that manufacturing growth induces productivity growth outside manufacturing, by absorbing idle or low productivity resources in other sectors. The growth of manufacturing itself is determined by the growth of demand, which must come from agriculture in the early stages of development, and from exports in the later stages. Kaldor’s original view⁷⁴ was that Britain’s growth rate was constrained by a shortage of labour, but he soon changed his mind in favour of the dynamic Harrod trade multiplier hypothesis of a slow rate of growth of exports in relation to the income elasticity of demand for imports, the ratio of which determines a country’s balance of payments constrained growth rate. Because fast growing ‘regions’ automatically become more competitive vis á vis slow growing regions, through the operation of the second ‘law’, Kaldor believed that growth will tend to be a cumulative disequilibrium process—or what Myrdal once called a ‘process of circular and cumulative causation’,—in which success breeds success and failure breeds failure. He articulated these ideas in several places, most notably in two lectures: his Inaugural Lecture at Cambridge in 1966,⁷⁵ and in the Frank Pierce Memorial Lectures at Cornell University in the same year.⁷⁶ Most of the debate concerning Kaldor’s growth laws has centred on Verdoorn’s Law and the existence of increasing returns. Kaldor drew inspiration for the theory from his early teacher, Allyn Young, and his neglected paper ‘Increasing Returns and Economic Progress’.⁷⁷ Young, in turn, derived his inspiration from Adam Smith’s famous dictum that productivity depends on the division of labour, and the division of labour depends on the size of the market. As the market expands, productivity increases, which in turn enlarges the size of the market. As Young wrote ‘change becomes progressive and propagates itself in a cumulative way’, provided demand and supply are elastic. Hence increasing returns is as much a macroeconomic phenomenon as a micro-phenomenon, which is related to the interaction between activities, and cannot be adequately discerned or measured by the observation of individual industries or plants. Kaldor was convinced by theoretical considerations and by his own research, and that of others, that manufacturing is different from agriculture and most service activities in its ability to generate increasing returns in the Young sense.

…

Notes

…

73. See A. P. Thirlwall (ed.), ‘Symposium on Kaldor’s Growth Laws’, Journal of Post-Keynesian Economics, Spring 1983.
74. See Causes of the Slow Rate of Economic Growth of the United Kingdom (CUP, 1966).
75. As note 74
76. Strategic Factors in Economic Development (Cornell University, Ithaca, New York, 1967).
77. Economic Journal, December 1928.

…

I don’t see much reference to short book Strategic Factors In Economic Development anywhere and I wasn’t even aware of the book till I reread this passage again recently. Must get it. Although according to this review, there’s nothing much in addition to Causes Of The Slow Rate Of Economic Growth Of The United Kingdom.

I have never been able to appreciate Kaldor’s earlier models, perhaps he tried to unsuccessfully build a stock-flow coherent model.

Also, in recent times, Post-Keynesians have come to the realisation that trade elasticities are endogenous not fixed parameters. To me that the most crucial aspect of circular and cumulative causation, although the Verdoorn Law plays a role too.