We estimate a dynamic panel data model to assess the relationship between dif ferent levels of instability-proxied by growth volatility and inflationand growth in Latin America from 1960 to 2011. Outlying observations could be mistakenly treated as thresholds or regime switch. Hence we use k-median clustering to mitigate the outlier problem and properly identify "scenarios" of instability. Our key findings are that while high inflation is harmful, low inflation is in fact positively related to growth. Volatility is also found to be significant and negative, but with no dif ferential ef fectbetween low and high levels-on growth.

For some time now, the difficulties in consolidating the MERCOSUR are no longer new. Today, economic and political instabilities appear to be the only sustainable and common facet in Latin American countries. Therefore, it is interesting to discuss with some evidence whether MERCOSUR is an institution that evolves consistently with its objectives despite the circumstances, or whether MERCOSUR is just a set of countries that like leaves pile up according to wind and luck. We addressed the periodization of MERCOSUR's lifetime based on a new approach. The main contribution of this paper is a precise delimitation of the stages in the evolution of the MERCOSUR from a method that reduces the risk of diverse types of biases. In fact, we showed evidence in favor of the first characterization, but the second prevails.

The paper proposes an interpretation of the «Argentine failure» based on development accounting and econometrical approaches frequently used in the current cross-country income differentials literature. The main results are as follows: the development process of Canada -in term of per capita GDP --moved away from that of Argentina around 1918, but there was a structural change in the determinants of aggregate productivity around 1935 that led Argentina to take a diverging path. Recovery -thanks to improved aggregate productivity --was not possible after 1940. The results support the idea that Argentina fell into a «staple trap», while Canada embarked on a successful path due to the adjacency and political proximity with a larger and complementary economy.Keywords: Argentine failure, staple trap, productivity, Argentina, Canada RESUMENEl artículo ofrece una interpretació n al «Fracaso Argentino» basado en el enfoque de contabilidad del desarrollo y la econometría, frecuentemente utilizados en la literatura actual que explica los diferenciales en el ingreso per capita. El proceso de desarrollo canadiense se alejó del argentino alrededor de 1918. Sin embargo, estas economías experimentaron un cambio estructural en los determinantes de la productividad agregada alrededor de 1935 que llevó a que también sean divergentes en términos tecnoló gicos. A partir de 1940, la recuperació n argentina no fue posible. Se ofrece soporte técnico a la idea que sostiene que Argentina cayó en una «staple trap», mientras Canadá ingresó en un sendero exitoso debido a su adyacencia y proximidad política con una economía mayor y complementaria.