Posted on September 13, 2020

Well, I feel like we’ve been here before. After employers were allowed back in late March to start deferring payment of their share of the Social Security payroll tax, now employees are allowed to do the same–though it will largely be employers who make the decision. The President’s executive action of August 8, allowing the employee tax deferral, kicked in on September 1, although in practice the start is delayed because the IRS guidance wasn’t released until August 28. Like the employer payroll tax deferral, the employee tax deferral lasts until the end of the calendar year. The employee tax deferred amounts must be paid back, in the form of increased withholding amounts, evenly from paychecks over the first four months of 2021; the employer-side deferred amounts don’t need to be paid back until the end of 2021 (half of the amount) and 2022 (the other half).

Estimates of federal revenue effects of legislative tax changes, or in this case executive actions, are most accurate when take-up rates aren’t involved–that is, when everyone is mandated to do something. With this executive order, the take-up rate–the proportion of employers who elect to undertake the deferral for their employees–will be critical for the effect on withholding, which we use to estimate withholding changes on a constant law basis so as to line up changes in withholding with wages and salaries.

The take-up rate for the employer-side payroll tax deferral from earlier in the year appears to have been quite anemic, and the current chatter is that we will also see anemic take-up of the employee payroll tax deferral. In our assessment of the take-up of the employer payroll taxes, from looking at financial statements, it appeared that large, public companies who elected to defer accounted for maybe 20 percent to 25 percent of national employment. When factoring in smaller employers who were less likely to defer, the percentage of total economy-wide employment was probably a bit lower.

Based on what we gather from payroll and other experts, we’re currently estimating that the executive action will reduce withholding by about 1.7 percent, not a large amount considering that employee-side payroll taxes make-up about 17.5 percent of total withholding tax receipts. For our estimates of withholding on a constant law basis, we’ll be adjusting the observed growth in withholding, which will include any deferrals, by that 1.7 percentage points, phased in over a period of weeks as employers make their decisions. Our estimate of 1.7 percent is based on the multiplication of three components (bolded below):

We should be able to monitor large employers to see what they are doing and adjust our estimate of 1.7 percent if necessary. It won’t be a state secret which firms are adopting the change for their employees because the workers can just look at their pay stubs. Richard Rubin of the Wall Street Journal has been doing a nice job of tracking whether employers are planning to defer the employee payroll tax or not. (Sorry, a subscription to the Journal is needed so I can’t link to his latest article.) So far large employers either seem to be electing not to defer or are still studying the issue, although Rubin’s sample is currently very small. In theory firms could allow employees to choose, but the costs of implementing that system could be prohibitive for many employers.

We have already been adjusting withholding growth by 4 percentage points to reflect the effects of legislation enacted in March, and our adjustment will now be (with a phase-in) about 5.7 percentage points. Because those law changes have all reduced withholding growth, we will take observed growth (which includes the effects of the law changes that have reduced withholding growth) and add 5.7 percentage points to measure growth on a constant law basis. Since we developed the estimate of 4 percentage points for the March legislation (see post of June 8), we have learned that the effects of the employer Social Security tax deferral have probably been a bit larger than we expected but the effects of the payroll tax credits for employee retention and sick and family leave have probably been a bit smaller. We infer that the 4 percentage point adjustment has been reasonably close to what has actually occurred because our measure of withholding growth adjusted for law changes has been lining up relatively closely with wage growth as measured in the monthly employment reports by the Bureau of Labor Statistics (see post of September 5). The introduction of law changes and executive actions that affect withholding do add uncertainty to our constant law measure of withholding growth.