Wealth Inequality and Savings

Derek Thompson has a piece at The Atlantic about wealth inequality and savings. In the piece, he rehashes some charts and figures from the latest Saez & Zucman (2014) offering that tracks US wealth inequality from 1913 to the present, and concludes that "despite the deep economic, psychological, and institutional pressures of low and stagnant wages, being wealthy starts with not spending money."

For the most part, Thompson relies upon this graph and particularly the recent dip in savings rate for the bottom 90% to conclude that "middle-class" savings plummeted and to animate his conclusion that increasing savings rates are the only way to increase the wealth share of the bottom 90%.

There are problems with this analysis.

First, Saez and Zucman define savings as the net increase in wealth "that is not due to changes in asset prices." This is a fine definition as far as things goes, but when combined with the dynamics of the housing bubble, it creates certain complications for the point Thompson is trying to make. Most middle-class wealth is tied up in housing and, because of the housing bubble, housing prices started running up in the 1990s. This meant that the wealth of these homeowning families, measured as total market value of the assets that they owned, went up over this period significantly.

It's quite probable that middle-class homeowners, upon observing that their wealth was increasing via capital gains in their homes, reduced the amount of income they were saving (behavior consistent with a well-known phenomenon known as the "wealth effect"). If you planned to put $50,000 of your income into savings over some period of time, but your home value unexpectedly shot up by $50,000 in that period, you might reasonably decide putting away that income is not necessary. Under the Saez-Zucman definition of savings, which excludes changes in asset prices, this shift registers as a decline in savings. But it's not clear it really is, or at minimum, it's the kind of decline in savings that has no net effect on wealth accumulation if everything goes according to plan. Of course, everything didn't go according to plan as the housing bubble burst and all of the excess wealth people thought they had built up in their home vanished.

Second, describing the bottom 90% as the "middle class" when referring to its savings rate is misleading. Those who have followed the parade of papers from Piketty, Saez, and Zucman know that they tend to define people in the bottom 50% as the lower class, people between the 50th and 90th percentiles as the middle class, and people in the top 10% as the upper class. From there, they note that when it comes to wealth levels and wealth shares, the bottom 90% is basically the same thing as the middle class because the bottom 50% owns basically zero net wealth. But the dynamic is different when it comes to savings rates. If the bottom 50% has a negative savings rate (meaning they are consuming more than they receive in capital and labor income), then that drags down the savings rate of the bottom 90%, including the "middle class" defined as the 50th to 90th percentiles. To figure out how the middle class is doing on savings, you really have to account for the differences in savings rates within the bottom 90%, which I suspect are significant.

Finally, it should be pointed out that, contrary to what Thompson says, evening out wealth inequality in this country does not necessarily requiring any change in savings rates. My preferred policy on handling wealth inequality is to levy a moderate progressive wealth tax and put all of the revenues from that tax into a sovereign wealth fund. This would reduce wealth levels at the top and, insofar as everyone could be deemed to have an equal share in the sovereign wealth fund, increase wealth levels at the bottom and middle. According to Zucman, a 2% net worth tax that fell only on the wealthiest 1% could deliver as much as $500 billion in revenue annually. I would also prefer assessing a lower wealth tax on smaller fortunes, but even just this 2% tax would allow for the construction of a $5 trillion sovereign wealth fund within a decade, delivering (with a 5% real return) $250 billion of revenue in investment returns alone in the 11th year (returns which would only get bigger as the fund grows from the wealth tax revenues). Those returns could go to all sorts of things, including a social dividend payment to all citizens, as in Alaska.

Comments