FLSA Data Research: How did we look for changes?
To investigate whether the proposed changes had had an effect before the December 1st deadline, we pulled quarterly salary information for more than 2 million respondents to PayScale’s compensation survey. This dataset contained almost 10,000 job titles, including everything from “Beef Cattle Nutritionist” to “Lead Ultrasonographer.” We grouped people based on whether their salaries were below the current threshold ($23,661), above the current threshold but below the proposed change (between $23,661 and $47,476), or above the proposed threshold (greater than $47,476).
Since we didn’t expect the changes to affect all workers (not many CEOs earn less than $47K, as an example), we restricted the data to only include job titles that already contained a substantial number workers with salaries between the old and new thresholds. Specifically, we only included jobs in which 20% or more of respondents earned between the current threshold and proposed change. This whittled our list of nearly 10,000 titles down to a much more manageable list of nearly 6,000 titles (dropping both “Beef Cattle Nutritionist” and “Lead Ultrasonographer,” unfortunately). Finally, to ensure that our estimates wouldn’t vary too wildly over time due to small sample size, we restricted the list to just job titles with at least 30 respondents per quarter.
For each of the jobs in the final list of 553 titles, we found the share of employees who would be eligible under the new threshold by calculating the percentage of workers earning less than $47,476 in each quarter from Q4 2014 through Q4 2016.
FLSA Data Analysis: What do the data show?
Given the buzz around the new rule in the media and in the compensation world, we expected to see the proportion of people earning less than $47K drop substantially in the two quarters after the announcement, and that’s exactly what we saw. To avoid tracking or paying overtime, companies have been increasing salaries past the proposed threshold to maintain employees’ exempt status. What we didn’t expect was the magnitude of the changes. In most jobs, the share of people earning less than $47K is basically constant until Q3 2016, when it drops suddenly. We can see this trend clearly in the graph below. There seems to be virtually no meaningful change relative to Q4 2014 until Q3 2016, when the line for nearly every job drops precipitously. For a typical job in this set, the proportion of workers earning less than the proposed threshold decreased by more than 18 percentage points between Q4 2015 and Q4 2016. The changes were consistent across jobs, as well. The share of respondents earning less than 47K decreased between Q4 2015 and Q4 2016 in 539 of our 553 titles.
The largest changes were in Policy Analysts (57 percentage point decrease), Staff Auditors (51 percentage point decrease), and Assistant Accountants (43 percentage point decrease). We also observed major changes in the time-intensive, white collar jobs that have been the focus the conversation about the proposed rule. For instance, the share of Middle School Teachers reporting salaries under $47K decreased from 72% in Q4 2015 to 47% in Q4 2016 (25 percentage points).
We should note that we cannot definitively say that the new threshold announcement caused these changes, but the patterns we’ve uncovered are striking. To quote the web comic XKCD, “correlation doesn’t imply causation, but it does waggle its eyebrows suggestively and gesture furtively while mouthing ‘look over there’.” Observing such a sharp drop in the share of people earning less than $47K after almost two years of negligible change is extremely suggestive.
FLSA Data Impact: What does it all mean?
Although the new overtime rule has been delayed by a federal judge in Texas and might face opposition in the Trump administration, our results indicate that it has already had a substantial effect on the salaries of many workers. The question now becomes: will these changes last? The economic theory of “wage-stickiness” suggests that it will. According to this idea, wages and salaries increase much more easily than they decrease because workers are unwilling to accept wage cuts. This makes it very difficult for the price of labor to decrease. Therefore, even if the new overtime exemption threshold is repealed, the salaries for the jobs it applies to might not decrease because pay has already ratcheted up. We will keep an eye on these jobs in the coming months, but we can already observe the wide-ranging effects of the rule change, even though it didn’t (and may never) go into effect.
For further reading: