I was a senior in high school in 1999, and it was pretty awesome — we played that Prince song a lot and didn't care about our grades. As it turned out, it was also the peak of real median household income in the United States for all major racial and ethnic groups, according to new data released today by the Census Bureau.
:format(webp):no_upscale()/cdn.vox-cdn.com/uploads/chorus_asset/file/4067414/Screen%20Shot%202015-09-16%20at%2010.35.18%20AM.png)
Yikes.
Now, you will see this chart a lot today, so it's worth keeping in mind that it comes with a lot of caveats.
- It is adjusted for inflation using the consumer price index. Most economists feel that a "chained" index like personal consumption expenditure deflator would be more accurate, and those indexes paint a more optimistic picture of income growth.
- By focusing on cash income, it excludes the value of things like employer-provided health benefits.
- You can't really measure the benefit of new technologies through this method. While I'm at work, I get iMessages from my baby's day care provider showing videos of him doing cute stuff, and I can share those videos with his grandparents. In 1999, *66 seemed like pretty cool technology. We've had some real improvements.
- Demographic change matters. Relative to 1999, we've become a less "middle-aged" society. We have a larger share of retired people than we used to, and we also have a larger share of 20-somethings. People in their 20s earn less than older people (and are less likely to be in two-income households than middle-aged people), and retired people have very low earnings.
All that said, data limitations have always applied to efforts to measure living standards. Those efforts used to show a clear upward trend. More recently, they have not. That's a change that should worry us, even if the real situation is likely less bleak than this chart makes it out to be.
