Something of a blow to Piketty's thesis

The more research that gets done into the details of Thomas Piketty's thesis (essentially, wealth concentration will leave us all as serfs again) the more there seem to be great gaping holes in it. For example, a central piece of the logic is that wealth will pile up, this will be inherited, and that wealth inequality will thus get worse over the generations. We're not convinced that such a bourgeois world would be a bad one but that thesis does depend upon the idea that inheritance concentrates wealth.

Which, apparently, it doesn't:

 Studies analysing the link between inheritance and wealth inequality have used different methods and data sources, ranging from simulated distributions to individual observations in surveys or data on wealth from tax records (e.g. Davies 1982, Wolff 2002, 2015, Boserup et al. 2016). Consensus has not yet been reached over the exact relationship. However, a recurrent result, which Wolff (2002) was the first to find, is that, perhaps surprisingly, inheritances tend to decrease wealth inequality.

Note that this isn't a new finding: it's just that Piketty assumed the opposite.

In a recent study (Elinder et al. 2016) we examine how inheritances affect wealth inequality using a new population-wide register database. The database contains detailed accounts on wealth and inheritances (including zeros) for all family and non-family heirs of every deceased person in Sweden during several years in the early 2000s. These rich data enable us to estimate the causal effect of inheritances on the distribution of wealth and, importantly, to also uncover the mechanisms underlying this effect. We are also able to study the distributional consequences of inheritance taxation and the effect of inheritances on wealth mobility.  

Our main results establish what previous studies have pointed to, namely that inheritances decrease the inequality of wealth.

The Gini coefficient decreases by 6%, a relatively large effect which is roughly in line with the wealth compression following the burst of the dotcom bubble in 2000 when stocks in internet companies, held mostly by the wealthy, lost their value.1

We also find that inheritances increase the absolute dispersion of wealth among heirs, measured as the difference in wealth between the heirs in the 25th and 75th percentiles of the distribution.

Isn't that fascinating? And more, the taxation of inheritances itself increases wealth inequality.

We observe the exact amount of inheritance taxes paid by each heir (some pay nothing), and when we examine how the tax payments affect the wealth distribution we find that the tax has a dis-equalising effect. That is, all else equal, the tax by itself tends to increase wealth inequality.

Of course, dependent upon what the money raised is spent upon the act of spending can reduce inequality again. But if inequality is the point and purpose of inheritance tax, which for many it is, there seems little point in increasing inequality by said taxing only to use the revenues to undo the effects just caused.

Our own longer term view has been that inheritance tax hasn't worked. The truly rich don't pay it, using trusts and lifetime gifts and so on. It's the less than plutocratic but still successful that do pay it. We've noted that old folk wisdom, clogs to clogs in three generations, and think that it has good predictive power. Even the inheritance of the grandest fortune cannot survive an inheritor truly determined to waste it and eventually, given the way genetics seems to work, one does always turn up to do such.

This might not draw nods of approval from those who would plan society but we've at least an urge to let people inherit as they may and leave the occasional existence of spendthrifts to deal with wealth concentration. There's very, very, few (in fact, other than those very few aristocratic families who held substantial urban land and kept it, we're not sure there are any) fortunes that have survived more than three generations.

The sands of time seem to deal well enough with this problem, why plan for it?