…is apparently less fluid than we tend to think. A really useful piece from the Economist updates us with the latest research from a variety of social scientists, and also–incredibly usefully–includes links to all the research it cites.
A second method relies on the chance overrepresentation of rare surnames in high- or low-status groups at some point in the past. If very few Britons are called Micklethwait, for example, and people with that name were disproportionately wealthy in 1800, then you can gauge long-run mobility by studying how long it takes the Micklethwait name to lose its wealth-predicting power. In a paper written by Mr Clark and Neil Cummins of Queens College, City University of New York, the authors use data from probate records of 19th-century estates to classify rare surnames into different wealth categories. They then use similar data to see how common each surname is in these categories in subsequent years. Again, some 70-80% of economic advantage seems to be transmitted from generation to generation.
It should by the way be mandatory for articles in newspapers and magazines published online to include links to the scientific papers to which they refer.