High and Rising Growth Continues in Economywide Tax Withholding and Wages

Posted on November 29, 2021

Summary

We have two new pieces of information to discuss, both of which suggest overall strength in the economy: in November, strong growth continued in the amounts of income and payroll taxes withheld from paychecks by employers and remitted to the Treasury (so-called tax withholding). Also, economywide wages and salaries as measured in the National Income and Product Accounts were revised up significantly last week for the second quarter (April-June) of 2021, and the strong growth is estimated to have become even higher in more recent months.

So, we measure that tax withholding–which is based largely on wages and salaries–was up by 17.6 percent in November compared to November of two years ago, before the pandemic. A measurement description if you’ll bear with me: that measure of withholding removes our estimated effect of tax law changes (which have been small in recent months) and standardizes the withholding amounts to reflect the number and makeup of business days across months. Without those adjustments, tax law changes that don’t reflect the underlying economy would be skewing the data, and withholding growth would also bounce around simply as a result of the calendar. Also, by looking back two years, we avoid the difficulty of measuring one-year growth rates off the low starting points from last year, and recent one-year growth rates are more susceptible to error measuring the effects of law changes on withholding, which were especially large last year. The amount of withholding is measured in nominal dollars, so it is not adjusted for inflation, which may become very important if the recent run-up in consumer prices continues. Also note that we can measure tax withholding for the full month of November even before the month is completely over, because we use representative periods from the month to measure growth, and we exclude the last couple of days of each month from that measurement (see our methodology).

That 17.6 percent two-year growth in withholding is very strong by historical standards (see chart below for growth over the past three years and the second chart below for a longer-term perspective). We haven’t seen growth rates this high since 2006, during the boom prior to the financial crisis bust that began in late 2007. Withholding growth has been moving up in an irregular fashion in recent months, and the overall move up indicates continued improvement in wages and salaries in the economy. We also believe that the strength in withholding indicates that wages and salaries have been moving up especially for higher-income individuals, but more on that later. Some of the wage increase probably reflects consumer price inflation in what can, if sustained, become a feedback loop, as higher wages per worker get partially passed onto consumers as higher prices, and higher prices generate higher wage demands by workers.

Tax withholding is largely based on wages and salaries in the economy, which in turn are measured by different sources, some available quickly and some with long lags. The two most current measures of wages and salaries are from the National Income and Product Accounts (the NIPAs, also known as the GDP accounts, released by the Bureau of Economic Analysis in the U.S. Department of Commerce) and from the monthly establishment survey released by the Bureau of Labor Statistics (BLS, in the U.S. Department of Labor) and that moves financial markets when released usually on the first Friday each month. The better measure is from the NIPAs, but it is available with more of a lag. The NIPA measure is better both because it includes additional wage sources, such as bonuses and income from stock option exercisings, and because the data source is administrative reports covering nearly the entire economy, rather than the more limited survey of establishments by BLS.

Wages and salaries in recent months as measured in the NIPAs have been growing a good bit faster than as measured by the BLS establishment survey (see chart above). Two-year growth in NIPA wages in October is currently measured at 12.9 percent, while the comparable measure from BLS (for private-sector workers only, the largest group available) was 8.5 percent. The NIPA wage growth was itself just revised up substantially for the second quarter of this year, and that carries into the data through October (see the dotted line in the chart above for how growth in NIPA wages looked before the revision). We were surprised by that upward revision (see post from last week).

Importantly, despite the upwardly revised growth in NIPA wages, withholding growth as we measure it still is significantly higher (again see the top chart for the recent period and the chart below for the past 15 years). Withholding normally grows a bit faster than wages in a growing economy, as higher wages per worker push more income into higher tax brackets and raise average tax rates. (For example, because income tax rates rise with income, if a worker receives a raise, the additional income is taxed at a higher rate than the average of all of the worker’s income before the raise.) We can see that tendency for withholding to grow slightly faster than wages over a long period of time when the economy is growing.

However, the excess of withholding growth over wage growth in recent months has been more than the normal amounts in a growing economy. That suggests that, at least in general, wages of higher-income workers have been growing faster than those of lower-wage individuals. In other words, average tax rates have increased both from economic growth overall as well as from the corresponding income growth being concentrated in higher-income workers who face higher tax rates and pay a disproportionate share of income taxes. Another possibility is that NIPA wages will be further revised up for the July-September quarter, like they were just revised up for the April-June quarter, to move them more into alignment with the withholding data. But we won’t get that NIPA wage revision for the third quarter of 2021 until February of next year.

The last point is that we will see if the strong withholding growth in November portends a strong employment report on Friday of this week, when the Bureau of Labor Statistics releases its establishment employment survey. Withholding growth has been bouncing around a bit in recent months at high levels, so it’s difficult to tell the signal from the noise on a month-to-month basis.