Debt Inequality v. Income Inequality

A lot of people are pushing the income inequality meme: that income inequality caused the financial crisis and subsequent (and continuing) economic depression.  A week or so ago, we had the IMF pushing it.  Robert Reich is currently peddling a book, Aftershock, which does the same.  And just today Linda Beale of Angry Bear and ataxingmatter queried "Inequality as a critical cause of the financial crisis? (by which she meant income inequality)."

My answer to Linda is no, income inequality was not a cause of the financial crisis, let alone a critical cause. 

The financial crisis was caused by debt-financed financial speculation gone bad.  While one is free to argue that wealth inequality arising from income inequality resulted in a lot of money seeking the high yields offered by financial speculation, if people hadn't used that money to borrow more money to leverage their bets, none of this would have happened.

The financial crisis caused the current economic downturn, called the "Great Recession" by Reich, which government hacks recently declared "over" back in June 2009 (although Bernanke and his cadre don't seem to believe it, either; you don't employ quantitative easing with a healthy economy).

Now we're at the point to ask whether income inequality matters.  Specifically, is the current economic downturn being prolonged by income inequality?

The facile answer would be yes; however, the correct answer is no.

Income inequality is a necessary but insufficient condition for suppressing demand (i.e. spending).  To suppress demand in the manner that we're currently witnessing requires income inequality coupled with debt inequality, which for the bottom 90% of American households is best measured by the debt-to-wage ratio.  If people did not have to spend their (increasingly shrinking) wages on servicing existing, overvalued debt, it would be available to spend on new purchases.

Unfortunately, the debt-to-wage ratio is something that is not tracked by anyone (at least to my knowledge).  I've seen the debt-to-income ratio and the savings rate, but both are typically tracked using BEA data, which, as I've documented previously, is highly suspect due to the fictional "imputations" it uses to inflate personal income, personal disposable income, and savings. IRS income data appears to be much more realiable (unlike BEA and Fed data, it is not subject to subsequent, self-serving revisions that can be detected only by reviewing old Census reports that captured the original data like a fly in amber), and Professor Emmanual Saez of UC Berkeley has a great IRS-based dataset available here (and lots of papers based on that dataset, in its various incarnations).  The Fed tracks debt data itself, but it relies on BEA data for income, disposable income, savings, etc.  Every few years, the Fed conducts its "Survey of Consumer Finances," which provides some information that is useful in connecting the dots between debt and income, but not necessarily between debt and wages.

A major limitation in the vast majority of the available economic data is the fact that it is expressed in the aggregate: the data that is made available for public consumption is for the ENTIRETY of the U.S. populace, and is not susceptible to a deeper dive to help understand what is really going on.  One of my favorite examples is the "median real income," which inherently measures the mid-point of all income, adjusted for inflation.  But what does the mid-point really tell us about income distribution when 50% of all income accrues to the top 10% of households (up from something like 35% thirty years ago)?  It should tell us that the so-called "stagnant" median real income masks a substantial decrease in real income for everybody not in the top 10%, but that's not what we hear.  We hear that everybody is just running in place.

To the extent that we ever see a parsing of the data, it depends entirely on what hypothesis the researcher is seeking to prove or disprove.  Thanks to Prof. Saez, we at least have a real drill down into the make up of the top decile of American households, at least as to income. 

Most income inequality researchers seem to focus on quintiles (five slices of 20%), but I don't find this meaningful for the simple fact that the Prof. Saez's data show that everybody below the top 5% is, in fact, a wage slave: at least 90% of the annual income for everybody below the 95th percentile of households comes from wages.  This is one reason for my emphasis on wages instead of income.  For 95% of American households, the ability to spend depends entirely upon the ability to earn.  If you lose your job, you can't (and won't) spend.  Period.

For those of us in the 95th percentile and above, well, we can keep on spending without a steady job because we've hoarded enough financial wealth that we can make money from our money and live on the fixed income doing so produces.

Now for the Data

First, I used the following Fed data (the D.3 table from the penultimate Flow of Funds report) to understand the amount of outstanding debt:

NOTE: I had to reconstruct the data for 1952-1974 using Fed data embedded in Census reports.

Second, I used the following Credit Suisse chart to estimate the relative amount of debt held by the top 10% of households versus everybody else:

FYI -- I believe that I've found the dataset Credit Suisse used to derive this chart at the Federal Reserve website, and I intend to check Credit Suisse's conclusions.  Regardless, however, a major question is what is meant by the term "income?"  The BEA measures income withouth capital gains.  The IRS provides measures with and without capital gains.  It is difficult to ascertain from the Fed survey data what is meant by "income," although it does seem to include income from all sources.  Moreover, the Fed slices and dices the survey data in truly odd ways that make it impossible to parse.  All that being said, I am assuming that "income" means income without capital gains.

Finally, I used Prof. Saez's data to calculate debt-to-wage ratios for 1988-2007 for the top decile of households versus everybody else, using IRS income data and calculating outstanding debt for the top decile using the Credit Suisse chart.

So, without further ado, here you go:

The final chart shows how the bottom 90% really piled on the debt in the last decade, while the top decile hovered around the historical mean.  That huge ramp in the ratio was no accident.

I'm happy to share my data sets, if anyone wants them.