How is the economy doing? Sounds like a simple question; it’s not. Many of the proxies we tend to use as a gauge — say, the Dow Jones industrial average, or GDP, or the top-line unemployment rate or the national debt — are nearly meaningless, or even misleading. At a minimum, they leave massive holes in the picture they paint about average Americans’ economic well-being. Those of us who care deeply about this question are constantly pouring over data sources and talking to people across the country, trying to find additional bits and pieces to fill out a more accurate portrait.
One of those data points is the personal savings rate. After people get paid, pay all their bills, and buy all their stuff, what’s left over? Sorry if that question makes you uncomfortable. That’s pretty common. For many Americans, the answer is stark: Less than zero. That indicates people are living fragile lives, just one piece of bad news away from a real personal disaster. Countless surveys have suggested this; only about half of Americans have enough savings to cover one-month’s living expenses.
Those are surveys, however. Government data hopefully provides a more accurate picture of what’s really going on — and a more macro picture, too. For many years, the Bureau of Economic Analysis has examined this question and come up with a single number to reflect our nation’s personal savings rate. Recently, the news has been pretty dark. A generation ago, the rate hovered around 10 percent. So for every $100 in income, our parents managed to sock away $10 for the future. In July, BEA data showed that American’s personal savings rate had fallen steadily from 2015 to 2018 — from 6 percent to around 3 percent, or barely above the rate of inflation. Essentially, that meant Americans weren’t saving anything.
Analysts seized on this data as a sign that America was headed for trouble.
Sam Ouliaris and Celine Rochon, in a working paper for the International Monetary Fund published recently, called out this trend, warning that recessions are often preceded by low personal savings rates.
“This issue has implications for the outlook for U.S. consumption and GDP growth, financial market stability, and external imbalances,” the authors wrote.
Then some amateur economist named Bob Sullivan picked up on it and issued a pretty stern warning in a story for Credit.com that was syndicated to places like USAToday.com.
“When consumers are spending money that they aren’t earning, this is a short-term boost that’s going to end one way or another,” I wrote. I’m now glad that I stuck to the usual economists’ trick of equivocating.
Because recently, BEA adjusted several years’ worth of data around the personal savings rate — a “tweak” based on new information from the IRS — and that tweak turned this storyline completely on its head. The savings rate is actually double what BEA thought. Today’s personal savings rate is actually 6.8%.
In other words, Americans have a decent amount of money left over at the end of the month — rather than nothing.
That’s quite a change. Both in data, and more important, in narrative.
What to make of this? How could something like this happen? I asked the two IMF economists what they thought of it, and they declined to comment, other than to say they will soon be issuing an update to their paper.
I felt compelled to do the same, so I called Brad Smith at BEA – his title is Chief, Compensation Section at the National Income and Wealth Division or BEA. He graciously spent a lot of time pouring over charts with me to explain what’s going on with the data.
BEA must take a massive amount of data points into account when it calculates the personal savings rate. You can look at the complex tables yourself, but I wouldn’t recommend it. After adding up Americans’ income sources — wages, interest income, rental income, and so on — it subtracts personal expenses (durable goods, services, etc.) and comes up with a tally. The leftovers is the savings rate.
All this data comes from various sources. It also covers various time periods. For some calculations, the data is relatively old, and BEA must extrapolate forward. That’s what happened here. Nearly 10-year-old numbers from the past were revised, and that had a big impact on today’s estimates.
Apparently, when people tell the government about their income, some people say one thing to the IRS and another thing to other researchers. Who knew? When the IRS looked at tax returns from 2008-2010, it found that a certain kind of taxpayer had often under-reported their income. So it adjusted that data, and sent it to BEA, which used this to make a new estimate for certain kinds of income in the 2013-2017 years. It’s still an estimate, but BEA feels this new extrapolation is far more accurate.
Note, this wasn’t just a garden-variety data revision. It was the largest such revision that observers can remember.
“We can confirm that this was the largest revision to personal saving and the personal saving rate since at least 2002,” Smith said.
So, does that mean Americans are doing much better than we thought? Not so fast.
According to Smith, and the tables he showed me, the vast majority of the increase comes from two places — non-farm proprietor’s income, and interest income. The proprietor’s income category dwarfs the interest income category, so that’s the real source of error. In some cases (2014 for example), business owners earned a full 10% more, or $132 billion more, than was previously believed.
In other words, people who own their own businesses are earning more income than we thought. So are people sitting on a pile of cash. So, good for them.
Does this mean the average American consumer is better off? Smith wouldn’t say.
“BEA doesn’t make that determination. It’s up to our users to interpret our data,” he said. But when I theorized that the increased income isn’t flowing to middle-class American laborers, he didn’t disagree.
There are legitimate reasons that income data regarding small business owners might be inconsistent. They routinely take tax deductions unavailable to standard filers, so there can be a big discrepancy between gross income and net income. That might be a source of data confusion.
But more to our point here: When trying to decide how the economy is doing, we are all playing a game of Blind Men and the Elephant. That doesn’t mean we shouldn’t keep trying, and I certainly will. But anyone who tells you with great certainty the economy is doing well, or doing poorly, is misleading you.
And me? Even though I appropriately equivocated, this story also serves as a correction of my earlier piece. It was based on the available data, but that data was wrong — at least for now. It seems that at least some Americans are doing a decent job saving money right now. Sorry about that. I’ll keep you posted on further revisions.
PS — Many professional economists love to scoff at the gig economy, tirelessly arguing that it’s all hype and isn’t showing up in government data as a significant phenomenon. Maybe this story is a hint as to why. More on that soon.