HOW MUCH DIFFERENCE DOES A LITTLE DIFFERENCE MAKE? SKEWED DISTRIBUTIONS, INNOVATIONS, AND MACROECONOMICS.

Gerald SILVERBERG (MERIT Maastricht University)

If the world were truly Gaussian (finite variance distributions) and the whole were simply the sum of its parts (the law of large numbers), large systems such as economies should be well behaved around their mean values. Instead we seem to live in a world where small events become amplified (technical innovations, fads and fashions) or where exceptionally large events occur with small but significant probabilites (stock market crashes, earthquakes, storms, wars). Statistically, both phenomena reveal themselves through highly-skewed distributions such as power laws, and may ultimately have similar explanations.

I will examine the empirical evidence for such skewness in the field of innovation studies and macroeconomic fluctuations and then examine some modeling approaches based on percolation theory and self-organized criticality. I will conclude by asking whether life in a highly skewed world is fundamentally different from life in a nice Gaussian one.