This work presents a robust empirical approach to dealing with the issue of the long run relationship between macroeconomic volatility, consumption behaviour and welfare for a large sample of countries. Differing from previous works, our empirical strategy is grounded on consumption and takes account of the role of persistence in consumption/income volatility. Our main conclusion is twofold: on one hand, we determine that aggregate volatility exerts, on average and ceteris paribus, a significant impact in deviating consumption from its smoothing path, producing aggregate extra saving, and in hampering future consumption prospects. This relationship holds across countries, including the poorest ones, and is becoming more significant in recent years. On the other hand, by deliberately proposing conservative estimates, we confirm Lucas' (1987) intuition about the low value of stabilisation policies, both across-countries and in the long-run. Our empirical evidence provides new insights into the long standing debate about the costs of fluctuations. First, it shows that the poorest countries hold a significant amount of "extra saving" too, because of economic fluctuations. Second, it underlies that, on average, uncertainty affects not only the volatility of consumption around its mean but the mean itself (Elbers and Gunning, 2003).