Declining Middle Class?

… Now the narrative is less about the declining middle class and more about a gilded age of the super-rich. The middle feels as if it has lost ground because of the extraordinary wealth accumulated by the very, very few. But that suggests that the pie is a fixed size, and that is clearly not the case.

What triggered the gilded age of the late 20th century? There are extensive arguments in the literature about the impact of tax cuts, about extraordinarily high executive compensation in top management, and more. But the recent research does not seem to support these two as major causes. The most compelling argument for this dramatic increase in income has been technological change. Those with the human capital to take advantage of the new technology gained, and continue to gain, at the expense of those who don't have it. This is the "winner take all" phenomenon. In the 1920s, cutting- edge technology meant electrification. Hence the spike in income at the very top level. The fact that inequality has increased in most of the richest countries is also consistent with this explanation. …

A few weeks ago I wrote a column that purported to address the connection between the rise in income inequality and the threats to the middle class. … But there were also complaints that I didn't really directly deal with the pervasive sense of declining fortune in the middle class. And for that I plead guilty. …

… Many people who responded to my earlier article pointed to data from the U.S. Census Bureau that show median household income adjusted for inflation increasing only 18% over the last 30 years. That is stagnation!

Again, however, aggregate data tell a different story–real income per person increased nearly 80% over the 30-year period. Now this could be consistent with the finding that the very top earners garnered all the gains, but it is more complicated than that. Fitzgerald of the Minneapolis Fed points out that the price index used by the Census Bureau overstates inflation and understates income gains. There has also been a dramatic change in the mix of household types–a decline in married-couple households and an increase in no-spouse households for both genders. This causes the Census data to understate significantly the median income gains for all categories of household.

Finally, the Census Bureau approach ignores several sources of income–benefits and transfers–that have grown significantly. When all of these pieces are out together, the data suggest that median Census income per person has risen by 50% over the past 30 years–not 18%.

There are two other pieces to this puzzle that deserve more extensive discussion. One is the observation that important changes have occurred in the composition and quality of the consumption of the median American consumer. Think back 30 years to the typical consumption bundle and ask how much better off you might feel with the current bundle and whether changes in median income reflect that.

The second piece of the puzzle is why the rise in inequality has been so dramatic–and why the skill premium has increased as much as it has. These issues are all connected….