Are Federal Workers Overpaid?
This CBO study is being touted as proof by conservatives that the answer to this question is a resounding “YES!” Going nearly as far is Megan McArdle, who nonetheless argues that “the CBO’s data suggest that we could probably get workers with a bachelor’s or lower for less money than we are now paying, and not suffer much decline in quality.”
A quick summation of the study’s results: on average, the federal government paid about 16% more in total compensation (wages + benefits) than a theoretically comparable private sector company would have paid after adjusting for thing like employee experience and education. On average, federal wages (ie, not including benefits) were only about 2% more than in the private sector, but were 20% higher for employees with no more than a high school diploma, and 23% lower for employees with graduate or professional degrees. The big difference came as a result of significantly higher benefit costs, which were on average almost 50% higher in the federal government, though benefits for employees with professional degrees were comparable in both the private sector and the federal government. Benefits for those with a high school diploma or less were nearly 75% greater than in the private sector on average. Moreover, though benefit costs in general make up the bulk of the difference in total compensation, most of that difference appears to come from the federal government’s offering of defined-benefit pension plans and the fact that it subsidizes retiree health insurance.
So does this mean that federal workers are overpaid, or that “we could probably get workers with a bachelor’s or lower for less money than we are now paying, and not suffer much decline in quality”? I’m not at all certain.*
To be sure, it’s clear that the market has no more than an indirect role in setting federal compensation levels. But the study actually explicitly rejects even McArdle’s relatively squishy claim, saying that an “assessment of how changes in the amount or composition of total compensation would affect the government’s ability to recruit and retain a qualified workforce is beyond the scope of this analysis.”
I think the CBO is right to make that disclaimer, which makes the actual utility of this study quite limited.
As the study notes, different employees can place vastly different values on compensation and wages, and the actual quality of an employee is ultimately unmeasurable in any meaningful way for purposes of a study like this. Moreover, because the lion’s share of the discrepancy comes from retirement benefits increasingly unavailable in the private sector, any discussion involving the caliber of the federal workforce is going to require a discussion as to the benefits of a workforce where workers have a strong incentive not constantly looking to change jobs the second something better comes along.
While it’s common to assume that job stability encourages laziness, the fact is that it also discourages experienced, quality, and yes, hardworking, employees from having a wandering eye. That means a workforce more dedicated to their employer, and less in need of constant retraining.
This is a dynamic that gets ignored in these types of discussions all too easily, I think. There may well be something to be said about the bad incentives created by job protections that make it difficult to impossible to fire a bad employee, but even with those bad incentives, it seems clear to me that the average new employee in any career path, no matter how much experience and training that employee possesses, is going to make boatloads more mistakes than the average employee with 5, 10, 15, 20 years experience working the same job or type of job for the same employer.
In other words, it may be possible, as McArdle suggests, for the federal government to obtain workers of similar experience and education for less money and, especially, less expensive retirement benefits. My point here is that it does not follow that this means we wouldn’t experience much of a dropoff in quality of government employee.
The other big discrepancy that seems to be at issue here is that wages and benefits for those with no more than a high school degree are both dramatically higher than in the private sector, with wages 21% higher and benefits an astounding 72% higher for this group.
Obviously, unionization is going to account for a sizable chunk of this particular discrepancy, since this group is surely the most heavily unionized in the federal government. But that still doesn’t answer the question of whether this group is overpaid or could be replaced with little or no dropoff in quality at significantly lower compensation packages. Again, though, it is not difficult to hypothesize a quality-focused, or at least “benefit-to-the taxpayers,” justification for the discrepancy.**
This is a segment of the population where, if we’re being honest, reliable employees are going to be harder to find such that a wage premium to obtain reliable employees may well be necessary.
It is also a segment of the population that, in the private sector, is treated as the most disposable, having to take a few steps back each time they have to switch jobs; if two people similarly situated in 1990 started at the same wage, worked equally hard and with equal competence, but one was laid off two or three times during that period and one was never laid off, the latter worker would be making significantly more than the one who was laid off a few times, even though their “observable traits” would be identical. So again we come back to the issue of the benefit a stable federal workforce might provide to the taxpayer.
Regardless, the most important problem with drawing the conclusion that McArdle and various conservatives draw from this study is that they are assuming that the federal government’s employment practices should be directly compared to the private sector. I submit that this is a fool’s errand, regardless of whether you view federal employee compensation as a moral issue (and I basically do not). I say this for the simple reason that it makes little sense for the federal government to operate like the average private business – taxpayers are not shareholders, and while there are no shortage of metrics we might choose to decide whether we have good government, the annual profit and loss statement isn’t one of them. To be sure, it may occasionally be a symptom of good government, but it is hardly the goal of good government; by contrast, profit maximization is (and should be!) an entirely appropriate and necessary goal of a private corporation.
Higher pay, and especially stronger benefits, almost assuredly make the federal government’s workforce more stable and more consistent. That means employees better capable of navigating the bureaucracy, who may well take more pride in their jobs, and who have more experience in their particular jobs with their particular employer.
The dynamic, globalized economy of the last few decades has, I assume, created a situation in which the private sector can little afford to place a premium on having a stable force of employees. The ability to quickly adapt is more important than having stable, well-trained, experienced, and loyal employees; shareholders, too, seem to now emphasize short term profitability over long-term security. In those circumstances, why worry about incentivizing a stable force of employees when your company’s product or services will probably be obsolete in two years, and you’ll either be headed towards bankruptcy, gobbled up and sold for parts, or in the process of completely retooling your blue collar workforce to adjust to new technological or competitive realities? But the federal government doesn’t have these concerns – it’s going to be providing basically the same products and services two years from now as it is today, no matter what happens.
None of this is to say that I think federal employees are definitively NOT overpaid or that reducing federal employee compensation would definitively hurt the quality of service provided by the federal government. I honestly don’t know one way or another. Instead, it is just to say that there are no shortage of alternative reasonable and plausible justifications for these much-ballyhooed differences in federal government compensation and private sector compensation, and it is of highly limited utility to compare those compensation levels exclusively and directly in order to answer the question “are federal employees overcompensated?”
* I confess up front that I hardly have the background in economics necessary to appropriately evaluate these claims, and so I have no intention of going any further than arguing that the conclusions which may be drawn from this study are limited.
**It is worth noting that only about 20% of federal workers fit into this category as compared to 40% of the workforce as a whole, and obviously this class of federal workers can be said to make substantially less on average than more educated federal workers. As a result, the extra discrepancies amongst this class are not as expensive to the taxpayer as they might otherwise be.