Munich Personal RePEc Archive

Entropy in the creation of knowledge: a candidate source of endogenous business cycles

Gomes, Orlando (2006): Entropy in the creation of knowledge: a candidate source of endogenous business cycles.


Download (176Kb) | Preview


Two sector growth models, with physical goods and human capital produced under distinct technologies, generally consider a process of knowledge obsolescence / depreciation that is similar to the depreciation process of physical goods. As a consequence, the long term rate of per capita growth of the main economic aggregates is constant over time. This rate can be endogenously determined (in endogenous growth models, where production is subject to constant returns) or it can be the result of exogenous forces, like technological progress or population dynamics (in neoclassical growth theory, where decreasing marginal returns prevail). In this paper, we introduce a new assumption about the generation of knowledge, which involves entropy, i.e., introducing additional knowledge to generate more knowledge becomes counterproductive after a given point. The new assumption is explored in scenarios of neoclassical and endogenous growth and it is able to justify endogenous fluctuations. Entropy in the creation of knowledge will imply that human capital does not grow steadily over time. Instead, cycles of various periodicities are observable for different degrees of entropy. Complete a-periodicity (chaos) is also found for particular values of an entropy parameter. This behaviour of the human capital variable spreads to the whole economy given that this input is used in the production of final goods and, thus, main economic aggregates time paths (i.e., the time paths of physical capital, consumption and output) will also evolve following a cyclical pattern. With this argument, we intend to give support to the view of endogenous business cycles in the growth process, which is alternative to the two mainstream views on business cycles: the RBC theory and the Keynesian interpretation.

MPRA is a RePEc service hosted by
the Munich University Library in Germany.