Income Inequality and the Consumption Gap

Kevin Hassett (aided by Aparna Mathur) is having a debate with the NYTimes on the income inequality issue. When Kevin writes:

This evidence has always struck us as counterintuitive because the National Income and Product Accounts data tell a story of robust growth in income and consumption.

One has to laugh as aggregates tell us nothing about the distribution of either income or consumption. So what is Kevin trying to say here? Is he arguing that a shift of income away from poor folks towards rich folks would have lowered consumption relative to income? Then he is making the argument that the marginal propensity to consume for rich folks is lower than it is for poor folks. But does Kevin reject the Ando-Modigliani life cycle model, which is essentially the same as the Friedman permanent income hypothesis? Stay tuned.

But Kevin also noted:

The Department of Labor’s Consumer Expenditure Survey provides detailed consumption data on a cross section of Americans; we can use this to estimate how much of our aggregate consumption went to each income group in recent years. These data reveal that the middle class has been doing pretty well for itself. Breaking the income distribution up into five quintiles, we tracked the consumption experience of the middle quintile (or middle class) in recent years … the real growth rate of consumption for the middle class between 2000 and 2005 was about 7.5 percent. So the welfare of the middle is improving, and unambiguously. There is nothing to debate. The strong economy is lifting all boats.

So consumption for one of the quintiles was 7.5% higher in real terms in 2005 than it was in 2000. A little perspective please. Real GDP rose by 12.5% over this period and the share of aggregate GDP that was consumed (thanks marcel) rose from 68.65% in 2000 to 70.19% in 2005. Consumption as a share of income may have increased because folks were silly enough to believe that George W. Bush was giving us our money back. Of course, these tax “cuts” were only tax shifts or deferrals, so the Barro-Ricardian Equivalence version of the life cycle model would have to wonder why folks were not saving for those future tax bills (more on this later). We should also note that the population rose by 5% over this period – so I have a question for Kevin. Is your 7.5% increase referring to aggregate consumption or consumption per capita? Let’s see how he started his discussion:

When George W. Bush took office in January 2001, the economy was producing roughly $10 trillion worth of output annually. This year, it looks like production will be in the neighborhood of $13.5 trillion. If we think of our nation’s GDP as analogous to an individual’s income, then that lucky fellow has seen his income increase over the past six years by about 35 percent. A person making $100,000 who saw the same increase in his income would now be making $135,000.

No, no, no, no. Aggregate nominal increases in GDP are not the same as the increase in real per capita GDP. Kevin Hassett is smart to know this so why does he open his piece with such an obvious misrepresentation?

We addressed the income inequality v. consumption inequality debate here:

Iacoviello’s model predicts that consumption inequality grows less than proportionally as income inequality grows, while wealth inequality grows more than proportionally. Examining the distribution of current consumption misses the point that living standards should be gauged across individual’s lifetimes. While the very wealthy have ample resources for old age retirement, the evidence suggests that the working poor may not – unless we retain the Social Security Trust Fund as a vehicle to supplement their retirement benefits.

What I’m trying to say is that it is lifetime consumption that matters. But then Kevin seems to be saying the same thing:

It also accounts for fluctuation of income over time. Consumption is generally smoothed over time by people to match their expected income. A person’s consumption provides a much more accurate read on how he’s doing than does his before-tax income.

Aha! Kevin does believe in the permanent income hypothesis after all! So let me see if I understand what he is saying. Folks in the lower quintile today are really rich people who have temporarily dropped out of the full time labor force, while the Bill Gates crowd is not consuming all of their income because they know they’ll take a few years off in the future. No – I don’t buy this explanation.

But as Kevin mentions before-tax v. after-tax income, maybe what he’s addressing is how we will reverse the Bush deficit disaster. It is reasonable to believe that someone’s future taxes will go up once we have a real leader in the White House. So is Kevin telling us that the middle class and the poor will keep their modest tax “cuts”, while the rich will have to endure a much more progressive tax code in the future? No – this is certainly not the Republican game plan. The few conservatives who are honest enough to admit taxes will have to increase (e.g., Bruce Bartlett and Greg Mankiw) are advocating a less progressive tax system. Maybe Kevin should tell us what type of tax system he would advocate. His spin here would make sense only if he believes the tax code will become more progressive. Is he advocating such a tax policy?

Update: Tyler Cowen and Greg Mankiw discuss the new book by Alan Reynolds who is also a consumption inequality buff.