Wealth Levels, Wealth Inequality, and the Great Recession [pdf]

However, just looking at the time scale over which the net worth changes described in the original article occurred, and their magnitudes, it doesn't seem likely that the effect they describe can be explained so easily by social factors. Hypothesizing that it's just a social change related to shrinking households leads to some interesting questions about what that really means in practical terms: Did the bottom 5% go from an average net worth of -$9K to -27K over the course of a decade because, on average, three individuals in the red would get together to form a combined household with pooled debts? Did the 25th percentile go from an average of $10k to an average of $3k because of parents getting divorced and kids moving out en masse? Meanwhile wealthy people's family structures remain static.

I think the data you link and the data in the article are reconcilable. Income and wealth are different things, after all. Perhaps wealthier folks' net worth dropped less precipitously during the Great Recession because they experienced nothing worse than a drop in the valuation of their primary residence. Meanwhile less wealthy folks suffered a greater loss because they were more likely to get foreclosed on, therefore suffering a loss of all the equity they had built up in their homes. This trend wouldn't show up as much in the income numbers because losing your house doesn't necessarily coincide with losing your job, certainly not the way things were playing out in 2008.

http://en.wikipedia.org/wiki/Wealth_inequality_in_the_United...

http://en.wikipedia.org/wiki/Affluence_in_the_United_States

1. If you own 10% equity in a home which declines in value 20%, you now have negative wealth. A 200% decline.

2. If you own 50% equity in a home which declines in value 20%, you have suffered a mere 40% decline.

Perhaps, but there is no hard data at the individual level because, as it happens, basically every comprehensive dataset on income or wealth relies on tax or census data, which are aggregated at the household level by design. So any empirical claim about individual-level data should be taken ex ante with a big grain of salt and evaluated carefully; especially in a heavily politicized field like economics where data is scarce and motivated reasoning reigns.

Take, for instance, the graph you cite [1], which cites Kitov and Kitov (2013) [2] along with the US census bureau as a source. The problem is, what that paper provides is not an empirical dataset but a microfounded model for predicting individual income --with all the problematic epistemological/overfitting caveats that they come with. Presenting that graph as an empirical fact and not as a debatable modeling exercise is quite misleading.

Besides, as others have pointed out, all the links you provide discuss income, not wealth (and no, their arguments/reasoning about income do not readily/trivially translate to wealth, as you commented below). If you are interested, check out the work of Cheryl Doss [3], who has conducted actual empirical work about the differences between household and individual equality in Ecuador, Ghana and India.

(PD: Oh wow even their claims about changing family structure in that blog [4] is disturbingly model-dependent. Much is made about the percentage of single-person households following an S-shaped curve, inflecting at 1968. However, they fit a sigmoid to what is a patently linear trend in the total number of single-person households beginning in 1950, and extrapolate back to 1900 despite having only two data points for that period. This wouldn't pass even a cursory peer-review, sorry.)

[1] http://1.bp.blogspot.com/-g3WZGpDibPM/Up3ZAPRtScI/AAAAAAAAJq...

[2] http://www.ecineq.org/ecineq_bari13/FILESxBari13/CR2/p177.pd...

[3] http://www.genderassetgap.org/sites/default/files/Doss%20-%2...

[4] http://politicalcalculations.blogspot.ca/2013/12/the-widows-...