My son worked with a state dept of enviromental quality as an intern for a year while getting his engineering degree. His first week on the job as an industrial permit engineer he worked through most of the backlog for the multi-media(air, water, and soil) applications in his section. His reward? The boss came in and told him to slow down. He knew right there was no way he would want to work in that environment. Private sector was next and he has had a good career.
Who watches the watchers? The Roberts talk was a good reminder of the need for accountability.
Again, as usual, no talk about limited terms for bureaucrats. No realistic change would improve effectiveness while reducing unaccountability as much as limiting “public service” to 8 years.
Was nice to read about FITs, and the criteria to score points. AI might get good enough to do a good job, but it would likely miss the deeper issue about what is the true narrative.
On news, there is an org called NewsGuard, which seems to mostly hassle conservatives, like
We need competing Rep and Dem audit agencies, each pointing out the bad stuff govt does, and the bad things done by the other Party. The general degradation of a stable bureaucracy full of CYA is even worse when combined with partisan enforcement of partisan regulations. Instead of asking for unrealistic non-partisans, as experts or DOJ agents or GAO auditors, we should be setting up govt systems so partisans of each Party act as checks, starting with publicity, against actions done or not done by the govt.
Trump could issue EO or guidance prohibiting career/career-conditional appointments and require only term appointments as the government is reorganized. Should send up a bill repealing the career appointment authority in title 5 USC. The savings would be enormous. And no impact on recruitment because fed employees would still be getting compensated much more than private industry. All the comparisons between federal and private pay are phony baloney and DOL won't release the raw data they are based on.
On the payroll thing, not really, it really depends on the specific job series / career field. Also you miss Federal pay isn't individualized hence you have to set a high floor if you want to attract talent, i.e. sure a GS-13 secretary makes vastly more than a receptionist at your local plumbing company but she makes vastly less than the secretary for Elon Musk. Likewise both the secretary and the doctor are GS-13s hence if you drop the pay because the secretary is "making too much", you are going to harm, and get even worse, doctors. The head of the National Weather Service makes around $200K, want to guess what the head of weather.com, AccuWeather, or FedEx makes? What lots of people don't grasp is Feds have normalized pay in a way private industry doesn't and that's a good thing.
In the aggregate though Feds make well under market wage in the positions where we expect high levels of expertise and competency.
PS: When the USG fired me for being competent after twenty years I was making $55/hr. My private sector peers were entry level $60/hr with most up near $90/hr. I nearly doubled my pay the next day. That isn't uncommon.
You are absolutely correct that one of the biggest problems with the federal pay system is that it does not take into account local variations in pay by occupation. Even the Federal Salary Council recognizes this:
“’The current pay comparison methodology used in the locality pay program ignores the fact that non-federal pay in a local labor market varies substantially between different occupational groups,’ the pay agent said in its October 2023 report. ‘As currently applied, locality payments in a local labor market may leave some mission-critical occupations significantly underpaid while overpaying others.’”
However, typically the federal government uses special pay rates for occupations deemed underpaid. For example, cybersecurity is often cited as an underpaid occupation, so the concerned agencies offer special rates:
“New employees with a bachelor degree hired into eligible positions can be paid a minimum salary of $76,156 under the supplemental rates, while those with a master’s degree could be offered a minimum of $88,250. And it gives agencies flexibility to offer increased rates for high-demand work roles or “exceptional qualifications.”
The highest end of the new pay system for cyber and STEM roles at the NSA and other defense intelligence agencies tops out at $183,500, the same as the salary cap under the new Special Salary Rate for IT employees being adopted by the Department of Veterans Affairs.”
By way of contrast, the average salary for a cybersecurity analyst in the United States is around $92,621 per year, with an average additional cash compensation of $5,955, with Entry-level candidates being offered $68,202 per year.
So, there really should not be a shortage of candidates for federal job openings. And indeed, the biggest complaint I ever heard from hiring officials was that if they publicly advertised a federal job opening they would be inundated with applications. So instead they chose to rely upon the plethora of alternative appointment authorities available in order to pick the person they preferred without having to pick from a certificate of eligibles.
But that is just one problem with the system. Others include that benefits are not included in federal-private comparisons, federal hiring authorities favor candidates with degrees for positions that do not require a degree (thus rendering CBO’s pay comparisons by education level irrelevant) and that the redacted BLS survey data is not publicly available. For my money though the biggest problem is that the federal position classification system against which private sector jobs are leveled is completely unmoored from any reality and every GS-13 position description talks about responsibilities that surpass Elon Musk’s and requires the ability to blow unicorn farts. But when you look at what is actually done by the employee encumbering the position, it is nearly always routine and insignificant memo generating administrivia. There just aren’t that many real scientists and mathematicians working in government.
The bottom line is that as of March this year, the average federal employee earned $106,382 in salary alone per year compared to the USA median 2023 salary of $48,060 for the median American. Americans made a 2023 median of $104,200 for computer and mathematical occupations, $99,220 for legal occupations, $79,050 for Business and financial operations occupations
I find it very difficult to believe that every federal employee is the equivalent of a Wall Street investment banker and is massively underpaid as the FEPCA system claims. It would be one thing if their mission was not to immiserate every ordinary American, but that is not the case and the sooner the federal bureaucracy is chopped down to size the better.
Because while the bottom wages have a floor (minimum wage) that is generally maybe a third the average GS wage, the top wouldn't hence you would see things like political appointees get multimillion dollar salaries eating up an even larger part of their agency budget and given no revenue restrictions, nothing would restrain them.
The Feds attempted a large multi agency experiment for about decade using flexible pay bands, what they unarguably found was universally it was a failure as every supervisor simply gave all their people by year three max pay regardless of performance whereas historically it takes about twenty years to max out via time in service via tenure hence was a net loss for the taxpayer for no gains. That program officially ended around 2021 as a failure though have good data.
Also from a pure employee relations egalitarian perspective, all Feds are equal. Whether you are a secretary or a doctor, you are first and foremost a bureaucrat hence no less important than any other peer of yours in the same pay class hence it decreases marginalization and encourages communication as you know what the shim you are talking to, regardless of job series or title, comparable level of responsibility. Plus likewise it's all public record, you know exactly what each of your fellow employees is making, they can just Google you hence you don't have the private sector pay politics here, i.e. the constant chip on your shoulder someone somewhere less productive than you might make a penny more.
It's true that flexible pay band compensation scheme did not live up to the unjustified hype or do anything to improve average levels of employee accountability or performance where they were implemented, but this was predictable from the very start from the type of work being done in banded positions in those agencies. Specifically, in many government jobs, there are rapidly diminishing returns to the input of an employee's additional competence or effort to the quality or quantity of desired deliverables when there is no extra value placed on those deliverables above a standard of adequacy. In such circumstances it doesn't make sense to provide incentives for better performance which it isn't actually possible for an employee with untapped potential to have even the opportunity to demonstrate.
On the other hand, the de facto practice in the banded schemes, while contrary to the hopes and intent of those who championed such schemes, did not produce any worse consequences than those caused by the GS scheme.
And when it really mattered that flattened pay scales started to misalign severely with market rates for workers in a particular specialty, such that one started to experience enduring and severe shortages of workers performing an indispensable function with the requisite marketable talents for those jobs - cybersecurity is the best recent example - then Congress always jumps in with some special 'extraordinary' kludge to 'temporarily' pay those particular workers the additional amount it takes to recruit and retain the minimal acceptable number of full time employees, and extends such temporary periods over and over indefinitely, while pretending they are still really "GS-scale employees" and the pay will go back to the standard amounts any day now. There are so many of these special plus-up provisions that the updated list of them all and their amounts is practically a government publication. There is no way to produce a fair comparison of compensation schemes with so many exceptions, upon which their functioning absolutely relied, under the simplifying assumption that those exceptions don't exist.
At any rate, where the bands were ended, it wasn't because they 'failed', but because the Democrats and the big government union organizations (but I repeat myself) were opposed to them in principle and as a matter of political and financial interest, and vowed to get rid of them as soon as they had the power to do so, and simply followed through on that promise, regardless of the meta-performance of the performance-based schemes.
> Congress always jumps in with some special 'extraordinary' kludge to 'temporarily' pay those particular workers the additional amount it takes to recruit and retain the minimal acceptable number of full time employees, and extends such temporary periods over and over indefinitely, while pretending they are still really "GS-scale employees" and the pay will go back to the standard amounts any day now
This is reminiscent of how in Edo period Japan samurai stipends (samurai having become public servants rather than warriors in all but name and outward attributes) were based on their family rank (which was fixed, hereditary, and next to impossible to improve) rather than ability. Daimyo often needed to put able lower-ranking samurai into nominally higher-ranking positions to handle business, but after the end of XVII c. they became very reluctant to promote samurai in family rank, to avoid upsetting the existing hierarchy. Some ways they created incentives for new hires was extra pay on top of their stipends, temporary stipend increases based on current position rather than family rank, or temporary, non-inheritable rank upgrades, but by far the most popular one was to give them actual decision powers far in excess of their family rank. High-ranking samurai were increasingly limited to the role of resolving conflicts and final approval, which gave legal power to the decisions bubbling up from the lower ranks, but they almost never initiated policy changes themselves. / 三谷博、維新史再考 (2017) p. 51 / This disconnect extended all the way to the top of the shogunate government, where small and middle daimyo unrelated to Tokugawa exercised actual powers while big daimyo who nominally far outranked them (notably this included both direct Tokugawa relatives and the rulers of provinces, such as Satsuma and Choshu, who sided against Tokugawa Ieyasu at Sekigahara) were excluded from policy-making and the shogun himself reduced to a figurehead. This disconnect became the driver of Bakumatsu and Meiji restoration, as high-rank daimyo clamored for their policy input to be heard by the government, and lower-rank samurai for their formal social position to match their actual contributions.
Seems to me you're arguing that government social engineering trumps market efficiency.
If *I* were playing fantasy games and trying to eliminate low-paying menial jobs, I'd do it by eliminating low-paying menial jobs, not by turning them into high-paying menial jobs.
This isn't a conversation about efficiency, it's about public policy coupled with the USG's, as an employer, HR strategy.
Also you aren't looking at the aggregate, i.e. doctors at a third the price, C levels at 1% the price.
Lastly people confuse Federal employees with State and Municipal employees who generally are vastly overpaid. They aren't the same thing. Federal employees get a lot of hate but realistically almost no one ever meets one in practice.
The most overpaid Federal employees are military enlisted but somehow everyone wants to pay them even more and I highly doubt when you put out your vitriol towards Federal employees you mean we should cut Private Smith's pay by 75%. I mean if you want to talk pure USG employee pay efficiency, conscripts are dirt cheap, should bring them back.
The federal government shouldn't be hiring doctors, health care should be a private industry. People should choose their doctors rather than the government choosing doctors for them. Very few private individuals would choose Dr. Fauci but plenty of government morons like Donald Trump would choose him.
Curious how many private sector doctors are willing to sleep in a tent and get shot at working 18x7 no overtime living off MREs in Iraq for year at time and the inability to quit or go on vacation. Those damn Federal government doctors.
Not saying you couldn't set the pay high but it's going to be exponential what you pay them now. You seem to forget on average the USG pays contractors two times what they pay in house when it comes to personnel, facilities, and services. When I left the USG the average rate we paid our contactors was $225/hr, we paid the Feds they replaced $115/hr factoring in benefits, facilities, etc costs.
I'll grant you that doctors working for the military in a legitimate defensive war should be paid a high salary. However, they're only a tiny portion of the doctors working for the government. The others are robbing the public.
Not at all, the overwhelming number of Federal doctors, not counting the ones in DoD, are either USPHS, VA, or the Bureau of Prisons all providing direct medical services, sans the latter, to those nominally assigned to them, i.e. USPHS serve as the doctors on NOAA Corp ships for example; the VA doctors, whom treat vets, are a huge cost savings over "care in the community" which is where the VA outsources vet care to the private sector.
I feel you are just repeating sound talking points with an ex to grind and have zero clue about Federal medical doctors. They aren't serving the public, you are confusing them with state and municipal doctors.
Fauci was not a currently licensed medical practitioner with prescription authority, he was a agency manager who happened to have a medical degree. I went to school for economics thirty years ago, that doesn't make me an economist.
Competing Dem and Rep audit agencies would be useless unless they were powerful government agencies, and if they were powerful government agencies, they would end up about as competitive as competing branches of government; when the government itself is at stake, they cover for each other.
When Republicans have a bad idea for a new law and can't pass legislation because the Democrats recognize the absurdity of the law, they combine it with a bad idea of the Democrats in a process called log-rolling to enact both bad ideas. This is illegal under state constitutions under single-subject rules but legal under the federal Constitution.
“End up” is always both subjective and unfalsifiable. The lousy agents we have now, after a couple hundred years, is where we’re at, but it’s not the end. Tho of course folks say we’ve ended up here, wherever they are when they speak. The closest we have to this model is the court system, with generally the opposing lawyers Not covering for each other.
I’m sure that for some good number of years, a partisan Reputation audit & internal regulation enforcement group would recommend a lot of govt bureaucrats be fired. New bureaucrats might be worse, but most often newer employees start out trying to do good work. Higher turnover will certainly improve most bureaucracies. Partisan audits to publicize govt failures will help, too.
Just like X community notes helps reduce misinformation, without censoring the words.
Before graduate school I worked at Lockheed Martin in Palo Alto and Sunnyvale. There on the Airborne Laser program I was steeped in the virtues of accountability and responsibility. This wasn’t necessarily a big part of the job description when I first arrived, but it came to be a huge part of the culture over the years that I was there. In fact, the day-to-day language and decision making processes are all about accountability and responsibility. Everyone knew what they were responsible for. It was part of your performance review. The entire system engineering process from concept development, critical design review, detailed design review to building and launching is a beautiful work of art and science.
After graduate school I worked at The Aerospace Corporation, an FFRDC overseeing the work of contractors like Lockheed, Boeing, Raytheon, etc. Here I was surprised that employees of Aerospace looked so negatively upon the defense contractors. My experience at Lockheed had been one of integrity, accountability and responsibility.
What I don’t know and couldn’t see in either organization is the ethical behavior of the executives at defense contractors. What holds them accountable? How much room is there for them to cheat tax payers?
"Nobody acts with the intention of doing away with accountability."
This sounds optimistic. I do believe that if banks have such armies of incompetent compliance teams, it is because the management believe their own accountability is diluted this way. The incompetents do not know what accountability is and believe that they are giving advice and guidance to those who are accountable (the ones who think they just shed their own accountability away).
Highly regulated industries end up acting as government bureaucracies do for the same reasons but I would not exclude intention on the part of some (a lot of) individuals.
Two Kling-themes that are in tension with the possibility of reaching the goal of improving accountability by means of clear job descriptions crafted specifically with that goal in mind.
Over 13 years ago you had a short post at Econlog, "The Job-Seekers Paradox" in which you made the point that many job-seekers want jobs that are as well-defined as possible, but the better defined a job is, the easier it is to automate (or outsource), so the less likely it is to continue to exist for long as stable employment for a local human. So, given that the capacity of automation technology keeps improving, by selection pressures, the labor market evolves to make the average job less and less well-defined.
The second Kling-theme - at least 15 years old from posts on Econlib - is to expound upon Garett Jones's observation that most workers don't make widgets anymore, instead they are producing organizational capital (or their work could be categorized as 'overhead'), and that it is inherently difficult to assess an individual employee's output in terms of the value added to the organization that could translate into marginal product that determines wages, and this gets increasingly difficult with ever-increasing complexity and specialization.
My point is that these trends, which are continuing to strengthen in significance because driven by technological developments, make it less likely that a sustainable local human job CAN be held accountable, as individual performance becomes harder to assess, and duties cannot sufficiently well-defined to provide the necessary basis for accountability-maximizing craftings of job descriptions.
I suspect that in addition to employees trying to evade accountability by adopting bureaucratic structures and procedures that diffuse responsibility, that this increased difficulty in holding workers accountability is also happening as a fully spontaneous consequence of the changing nature of the work itself, as a kind of emergent phenomenon that appears as the capabilities of automation technology pass a certain level.
One prediction is that with that capability looking to suddenly grow fast with AI, that the substitution will occur disproportionately precisely to the jobs easiest to hold accountable, and thus the jobs that most humans will do will rapidly become inherently resistant to any efforts to hold them accountable by defining standards of performance.
So, in the near future, every job will become increasingly like a government job. Have a nice day!
I've a general bureaucracy theory which I bet someone smarter has already invented, and has been named after someone else.
1. Start a small company. One person handles all HR and office functions: tracking vacations, sick time, pay rates, new hires, old fires, stocking the break room. Probably has a real job too.
2. More people are hired. At some point, the HR person is working 50 hours a week, so the company hires an HR clerk. Now they together work 60 hours a week.
3. Wouldn't do to have the new full-time clerk only work half time, or loan him out to other departments, so the HR boss finds new work. Organize a Christmas party, river and trees outings for morale, monthly lunches and go kart afternoons.
4. And so on. Company eventually gets to 100 employees and 3 in HR.
5. Then trouble. Market changes, new competition, economic downturn. Lay off production and design employees, but HR? Not on your life! Employees have come to expect go kart afternoons, and too much cutback leads to industry rumors of not doing so well.
6. But markets force changes, if businesses won't make them voluntarily.
And 6 is where government goes off the rails. They are monopolies.
The US created a tea importation board in 1897 to verify "that the tea should be rejected if it was unfit" which bureaucrats of course expanded to include tea tasting. It was finally abolished in 2023, 27 years after its budget had been zeroed out and was no longer taste-testing tea.
There's also the possibility of structuring government agencies as corporations, or farming out government operations to private bidders. There are problems with this approach, but are they worse than the problems associated with unaccountable agencies? Prior to the 20th century, the state often compensated public servants by facilitative payments and bounties. To cut wasteful spending and improve public services, perhaps we should consider partial reversions to the old system: https://www.amazon.com/Against-Profit-Motive-Revolution-Government/dp/0300194757
Arnold Kling writes about position descriptions today. He advocates using the PD for “aligning authority with accountability.” He observes that “In government, the tendency is to minimize the authority of any individual. Instead, a typical action is subject to a complex approval process. This makes it very difficult to assign accountability.”
and in the performance appraisal system under chapter 43 of title 5, United States Code (https://uscode.house.gov/browse/prelim@title5/part3/subpartC/chapter43&edition=prelim ). Further, the new political appointee should be aware of the organizational alignment of authority and accountability supposedly provided under the strategic and performance plans originally required by the Governmnment Performance and Results Act (see 5 U.S.C. 306 and 31 U.S.C. 1115) with which individual performance appraisal plans are supposed to be in alignment. In fact, it would probably behoove anyone who is going to be supporting or helping the Department of Government Efficiency to become well acquainted with title 5, United States Code generally: success will likely turn on mastery of the bureaucratic procedures authorized there.
However, eager reformers should not be intimidated by the morass, but rather take heart in knowing that government efficiency is embedded in the statutory merit system principles: 5 U.S.C. 2301(b) provides in part:
“(4) All employees should maintain high standards of integrity, conduct, and concern for the public interest.
(5) The Federal work force should be used efficiently and effectively.
(6) Employees should be retained on the basis of the adequacy of their performance, inadequate performance should be corrected, and employees should be separated who cannot or will not improve their performance to meet required standards.”
And moreover, it is a prohibited personnel practice under 5 U.S.C. 2302 for any manager or supervisor, career or not, to take personnel actions to prevent an employee from disclosing waste:
“8) take or fail to take, or threaten to take or fail to take, a personnel action with respect to any employee or applicant for employment because of-
(A) any disclosure of information by an employee or applicant which the employee or applicant reasonably believes evidences-
(i) any violation of any law, rule, or regulation, or
(ii) gross mismanagement, a gross waste of funds, an abuse of authority, or a substantial and specific danger to public health or safety...”
The first thing to know about position descriptions in the central government is that they are the principal document upon which pay for the position will be determined. Therefore, the document has to make all sorts of outlandish claims about importance and difficulty in order to support the pay grade of the position’s supervisor. 5 U.S.C. 5104 (https://uscode.house.gov/view.xhtml?req=granuleid:USC-prelim-title5-section5104&num=0&edition=prelim ) sets out the general criteria for grading positions.
At the top of the General Schedule pay system under which most central government employees are paid, is the GS-15 grade. 5 U.S.C. 5104 states:
“(15) Grade GS–15 includes those classes of positions the duties of which are-
(A) to perform, under general administrative direction, with very wide latitude for the exercise of independent judgment, work of outstanding difficulty and responsibility along special technical, supervisory, or administrative lines which has demonstrated leadership and exceptional attainments;
(B) to serve as head of a major organization within a bureau involving work of comparable level;
(C) to plan and direct or to plan and execute specialized programs of marked difficulty, responsibility, and national significance, along professional, scientific, technical, administrative, fiscal, or other lines, requiring extended training and experience which has demonstrated leadership and unusual attainments in professional, scientific, or technical research, practice, or administration, or in administrative, fiscal, or other specialized activities; or
(D) to perform consulting or other professional, scientific, technical, administrative, fiscal, or other specialized work of equal importance, difficulty, and responsibility, and requiring comparable qualifications.”
As of March 2024, the central government’s Office of Personnel Management reports on its FEDSCOPE statistical system that of the central government’s 2,278,730 total employees, some 71,780 individuals, or about 3% of the total workforce, encumbered positions graded at the GS-15 level. The first lesson, then should obviously be that the federal pay and classification systems and everything about them are a farce. It is probably much worse on the military side where one can’t swing a dead cat without hitting a flag officer and the ratio of enlisted to officer is wholly absurd. Total central government payroll obligations for 2023 amounted to $686,716 million or over 10% of net federal outlays of $6,134,672 million for that year.
In the central government personnel system, the performance appraisal system is where authority and accountability have a putative alignment. Yet rarely will the individual performance appraisal plan align with the position description of the incumbent in terms of all the exceptional and outstanding duties assigned to the position. In fact, it all dissolves into a hopeless muddle. And for a large share of central government employees, the performance appraisal output measures are negotiated in union contracts. A bit of sunlight might disinfect this sordid state of affairs, and DOGE might well consider requiring agencies to post all of their union contracts alongside their budget documents on their agency websites.
Because position descriptions are not aligned with performance appraisal plans, government reformers have tried to impose top down accountability with things like the oft amended Government Performance and Results Act which list an agency’s work plan and the outcomes for which it intends to hold itself accountable. These documents too are risible but they are a good place to start in figuring out why statutory programs are being neglected in favor of nebulous feel good performance measures for which accountability is largely impossible.
Plus government arguably has extra failure modes. Like how do we limit authority while adding accountability? A strongman or Spoils system president has lots of accountability in principle, but little in practice. Process accountability, vs pure outcome accountability, is expected not only to foster reliability but to limit corruption and power grabs.
I guess it comes back to big public choice dilemmas like lack of competition (only one federal gov), complexity (fuzzy relationship between policy and outcome), voter incentives... A corrupt CEO ultimately faces shareholders. And i feel like the libertarian answer of limited government kind of begs the question.
For a long time, my little charter project had a motto something like this:
When authority exceeds accountability, you have corruption.
When accountability exceeds authority, you have scapegoats.
My son worked with a state dept of enviromental quality as an intern for a year while getting his engineering degree. His first week on the job as an industrial permit engineer he worked through most of the backlog for the multi-media(air, water, and soil) applications in his section. His reward? The boss came in and told him to slow down. He knew right there was no way he would want to work in that environment. Private sector was next and he has had a good career.
Who watches the watchers? The Roberts talk was a good reminder of the need for accountability.
Again, as usual, no talk about limited terms for bureaucrats. No realistic change would improve effectiveness while reducing unaccountability as much as limiting “public service” to 8 years.
Was nice to read about FITs, and the criteria to score points. AI might get good enough to do a good job, but it would likely miss the deeper issue about what is the true narrative.
On news, there is an org called NewsGuard, which seems to mostly hassle conservatives, like
https://www.powerlineblog.com/archives/2024/11/who-will-guard-newsguard.php
We need competing Rep and Dem audit agencies, each pointing out the bad stuff govt does, and the bad things done by the other Party. The general degradation of a stable bureaucracy full of CYA is even worse when combined with partisan enforcement of partisan regulations. Instead of asking for unrealistic non-partisans, as experts or DOJ agents or GAO auditors, we should be setting up govt systems so partisans of each Party act as checks, starting with publicity, against actions done or not done by the govt.
Trump could issue EO or guidance prohibiting career/career-conditional appointments and require only term appointments as the government is reorganized. Should send up a bill repealing the career appointment authority in title 5 USC. The savings would be enormous. And no impact on recruitment because fed employees would still be getting compensated much more than private industry. All the comparisons between federal and private pay are phony baloney and DOL won't release the raw data they are based on.
On the payroll thing, not really, it really depends on the specific job series / career field. Also you miss Federal pay isn't individualized hence you have to set a high floor if you want to attract talent, i.e. sure a GS-13 secretary makes vastly more than a receptionist at your local plumbing company but she makes vastly less than the secretary for Elon Musk. Likewise both the secretary and the doctor are GS-13s hence if you drop the pay because the secretary is "making too much", you are going to harm, and get even worse, doctors. The head of the National Weather Service makes around $200K, want to guess what the head of weather.com, AccuWeather, or FedEx makes? What lots of people don't grasp is Feds have normalized pay in a way private industry doesn't and that's a good thing.
In the aggregate though Feds make well under market wage in the positions where we expect high levels of expertise and competency.
PS: When the USG fired me for being competent after twenty years I was making $55/hr. My private sector peers were entry level $60/hr with most up near $90/hr. I nearly doubled my pay the next day. That isn't uncommon.
You are absolutely correct that one of the biggest problems with the federal pay system is that it does not take into account local variations in pay by occupation. Even the Federal Salary Council recognizes this:
“’The current pay comparison methodology used in the locality pay program ignores the fact that non-federal pay in a local labor market varies substantially between different occupational groups,’ the pay agent said in its October 2023 report. ‘As currently applied, locality payments in a local labor market may leave some mission-critical occupations significantly underpaid while overpaying others.’”
(https://federalnewsnetwork.com/pay/2023/10/pay-agent-renews-calls-for-major-legislative-reforms-to-federal-pay/ )
However, typically the federal government uses special pay rates for occupations deemed underpaid. For example, cybersecurity is often cited as an underpaid occupation, so the concerned agencies offer special rates:
“New employees with a bachelor degree hired into eligible positions can be paid a minimum salary of $76,156 under the supplemental rates, while those with a master’s degree could be offered a minimum of $88,250. And it gives agencies flexibility to offer increased rates for high-demand work roles or “exceptional qualifications.”
The highest end of the new pay system for cyber and STEM roles at the NSA and other defense intelligence agencies tops out at $183,500, the same as the salary cap under the new Special Salary Rate for IT employees being adopted by the Department of Veterans Affairs.”
(https://federalnewsnetwork.com/pay/2023/08/pentagon-approves-higher-cyber-pay-for-nsa-other-defense-intelligence-agencies/ )
By way of contrast, the average salary for a cybersecurity analyst in the United States is around $92,621 per year, with an average additional cash compensation of $5,955, with Entry-level candidates being offered $68,202 per year.
So, there really should not be a shortage of candidates for federal job openings. And indeed, the biggest complaint I ever heard from hiring officials was that if they publicly advertised a federal job opening they would be inundated with applications. So instead they chose to rely upon the plethora of alternative appointment authorities available in order to pick the person they preferred without having to pick from a certificate of eligibles.
But that is just one problem with the system. Others include that benefits are not included in federal-private comparisons, federal hiring authorities favor candidates with degrees for positions that do not require a degree (thus rendering CBO’s pay comparisons by education level irrelevant) and that the redacted BLS survey data is not publicly available. For my money though the biggest problem is that the federal position classification system against which private sector jobs are leveled is completely unmoored from any reality and every GS-13 position description talks about responsibilities that surpass Elon Musk’s and requires the ability to blow unicorn farts. But when you look at what is actually done by the employee encumbering the position, it is nearly always routine and insignificant memo generating administrivia. There just aren’t that many real scientists and mathematicians working in government.
The bottom line is that as of March this year, the average federal employee earned $106,382 in salary alone per year compared to the USA median 2023 salary of $48,060 for the median American. Americans made a 2023 median of $104,200 for computer and mathematical occupations, $99,220 for legal occupations, $79,050 for Business and financial operations occupations
and $80,820 Healthcare practitioners and technical occupations. And the federal benefits are much better and even the FERS employees get a defined benefit retirement. See: https://www.bls.gov/emp/tables/emp-by-major-occupational-group.htm and compare at OPM’s FEDSCOPE app https://www.fedscope.opm.gov/
I find it very difficult to believe that every federal employee is the equivalent of a Wall Street investment banker and is massively underpaid as the FEPCA system claims. It would be one thing if their mission was not to immiserate every ordinary American, but that is not the case and the sooner the federal bureaucracy is chopped down to size the better.
Why on earth is it a good thing?
Because while the bottom wages have a floor (minimum wage) that is generally maybe a third the average GS wage, the top wouldn't hence you would see things like political appointees get multimillion dollar salaries eating up an even larger part of their agency budget and given no revenue restrictions, nothing would restrain them.
The Feds attempted a large multi agency experiment for about decade using flexible pay bands, what they unarguably found was universally it was a failure as every supervisor simply gave all their people by year three max pay regardless of performance whereas historically it takes about twenty years to max out via time in service via tenure hence was a net loss for the taxpayer for no gains. That program officially ended around 2021 as a failure though have good data.
Also from a pure employee relations egalitarian perspective, all Feds are equal. Whether you are a secretary or a doctor, you are first and foremost a bureaucrat hence no less important than any other peer of yours in the same pay class hence it decreases marginalization and encourages communication as you know what the shim you are talking to, regardless of job series or title, comparable level of responsibility. Plus likewise it's all public record, you know exactly what each of your fellow employees is making, they can just Google you hence you don't have the private sector pay politics here, i.e. the constant chip on your shoulder someone somewhere less productive than you might make a penny more.
It's true that flexible pay band compensation scheme did not live up to the unjustified hype or do anything to improve average levels of employee accountability or performance where they were implemented, but this was predictable from the very start from the type of work being done in banded positions in those agencies. Specifically, in many government jobs, there are rapidly diminishing returns to the input of an employee's additional competence or effort to the quality or quantity of desired deliverables when there is no extra value placed on those deliverables above a standard of adequacy. In such circumstances it doesn't make sense to provide incentives for better performance which it isn't actually possible for an employee with untapped potential to have even the opportunity to demonstrate.
On the other hand, the de facto practice in the banded schemes, while contrary to the hopes and intent of those who championed such schemes, did not produce any worse consequences than those caused by the GS scheme.
And when it really mattered that flattened pay scales started to misalign severely with market rates for workers in a particular specialty, such that one started to experience enduring and severe shortages of workers performing an indispensable function with the requisite marketable talents for those jobs - cybersecurity is the best recent example - then Congress always jumps in with some special 'extraordinary' kludge to 'temporarily' pay those particular workers the additional amount it takes to recruit and retain the minimal acceptable number of full time employees, and extends such temporary periods over and over indefinitely, while pretending they are still really "GS-scale employees" and the pay will go back to the standard amounts any day now. There are so many of these special plus-up provisions that the updated list of them all and their amounts is practically a government publication. There is no way to produce a fair comparison of compensation schemes with so many exceptions, upon which their functioning absolutely relied, under the simplifying assumption that those exceptions don't exist.
At any rate, where the bands were ended, it wasn't because they 'failed', but because the Democrats and the big government union organizations (but I repeat myself) were opposed to them in principle and as a matter of political and financial interest, and vowed to get rid of them as soon as they had the power to do so, and simply followed through on that promise, regardless of the meta-performance of the performance-based schemes.
> Congress always jumps in with some special 'extraordinary' kludge to 'temporarily' pay those particular workers the additional amount it takes to recruit and retain the minimal acceptable number of full time employees, and extends such temporary periods over and over indefinitely, while pretending they are still really "GS-scale employees" and the pay will go back to the standard amounts any day now
This is reminiscent of how in Edo period Japan samurai stipends (samurai having become public servants rather than warriors in all but name and outward attributes) were based on their family rank (which was fixed, hereditary, and next to impossible to improve) rather than ability. Daimyo often needed to put able lower-ranking samurai into nominally higher-ranking positions to handle business, but after the end of XVII c. they became very reluctant to promote samurai in family rank, to avoid upsetting the existing hierarchy. Some ways they created incentives for new hires was extra pay on top of their stipends, temporary stipend increases based on current position rather than family rank, or temporary, non-inheritable rank upgrades, but by far the most popular one was to give them actual decision powers far in excess of their family rank. High-ranking samurai were increasingly limited to the role of resolving conflicts and final approval, which gave legal power to the decisions bubbling up from the lower ranks, but they almost never initiated policy changes themselves. / 三谷博、維新史再考 (2017) p. 51 / This disconnect extended all the way to the top of the shogunate government, where small and middle daimyo unrelated to Tokugawa exercised actual powers while big daimyo who nominally far outranked them (notably this included both direct Tokugawa relatives and the rulers of provinces, such as Satsuma and Choshu, who sided against Tokugawa Ieyasu at Sekigahara) were excluded from policy-making and the shogun himself reduced to a figurehead. This disconnect became the driver of Bakumatsu and Meiji restoration, as high-rank daimyo clamored for their policy input to be heard by the government, and lower-rank samurai for their formal social position to match their actual contributions.
Seems to me you're arguing that government social engineering trumps market efficiency.
If *I* were playing fantasy games and trying to eliminate low-paying menial jobs, I'd do it by eliminating low-paying menial jobs, not by turning them into high-paying menial jobs.
This isn't a conversation about efficiency, it's about public policy coupled with the USG's, as an employer, HR strategy.
Also you aren't looking at the aggregate, i.e. doctors at a third the price, C levels at 1% the price.
Lastly people confuse Federal employees with State and Municipal employees who generally are vastly overpaid. They aren't the same thing. Federal employees get a lot of hate but realistically almost no one ever meets one in practice.
The most overpaid Federal employees are military enlisted but somehow everyone wants to pay them even more and I highly doubt when you put out your vitriol towards Federal employees you mean we should cut Private Smith's pay by 75%. I mean if you want to talk pure USG employee pay efficiency, conscripts are dirt cheap, should bring them back.
The federal government shouldn't be hiring doctors, health care should be a private industry. People should choose their doctors rather than the government choosing doctors for them. Very few private individuals would choose Dr. Fauci but plenty of government morons like Donald Trump would choose him.
Curious how many private sector doctors are willing to sleep in a tent and get shot at working 18x7 no overtime living off MREs in Iraq for year at time and the inability to quit or go on vacation. Those damn Federal government doctors.
Not saying you couldn't set the pay high but it's going to be exponential what you pay them now. You seem to forget on average the USG pays contractors two times what they pay in house when it comes to personnel, facilities, and services. When I left the USG the average rate we paid our contactors was $225/hr, we paid the Feds they replaced $115/hr factoring in benefits, facilities, etc costs.
I'll grant you that doctors working for the military in a legitimate defensive war should be paid a high salary. However, they're only a tiny portion of the doctors working for the government. The others are robbing the public.
Not at all, the overwhelming number of Federal doctors, not counting the ones in DoD, are either USPHS, VA, or the Bureau of Prisons all providing direct medical services, sans the latter, to those nominally assigned to them, i.e. USPHS serve as the doctors on NOAA Corp ships for example; the VA doctors, whom treat vets, are a huge cost savings over "care in the community" which is where the VA outsources vet care to the private sector.
I feel you are just repeating sound talking points with an ex to grind and have zero clue about Federal medical doctors. They aren't serving the public, you are confusing them with state and municipal doctors.
Fauci was not a currently licensed medical practitioner with prescription authority, he was a agency manager who happened to have a medical degree. I went to school for economics thirty years ago, that doesn't make me an economist.
Competing Dem and Rep audit agencies would be useless unless they were powerful government agencies, and if they were powerful government agencies, they would end up about as competitive as competing branches of government; when the government itself is at stake, they cover for each other.
When Republicans have a bad idea for a new law and can't pass legislation because the Democrats recognize the absurdity of the law, they combine it with a bad idea of the Democrats in a process called log-rolling to enact both bad ideas. This is illegal under state constitutions under single-subject rules but legal under the federal Constitution.
“End up” is always both subjective and unfalsifiable. The lousy agents we have now, after a couple hundred years, is where we’re at, but it’s not the end. Tho of course folks say we’ve ended up here, wherever they are when they speak. The closest we have to this model is the court system, with generally the opposing lawyers Not covering for each other.
I’m sure that for some good number of years, a partisan Reputation audit & internal regulation enforcement group would recommend a lot of govt bureaucrats be fired. New bureaucrats might be worse, but most often newer employees start out trying to do good work. Higher turnover will certainly improve most bureaucracies. Partisan audits to publicize govt failures will help, too.
Just like X community notes helps reduce misinformation, without censoring the words.
Before graduate school I worked at Lockheed Martin in Palo Alto and Sunnyvale. There on the Airborne Laser program I was steeped in the virtues of accountability and responsibility. This wasn’t necessarily a big part of the job description when I first arrived, but it came to be a huge part of the culture over the years that I was there. In fact, the day-to-day language and decision making processes are all about accountability and responsibility. Everyone knew what they were responsible for. It was part of your performance review. The entire system engineering process from concept development, critical design review, detailed design review to building and launching is a beautiful work of art and science.
After graduate school I worked at The Aerospace Corporation, an FFRDC overseeing the work of contractors like Lockheed, Boeing, Raytheon, etc. Here I was surprised that employees of Aerospace looked so negatively upon the defense contractors. My experience at Lockheed had been one of integrity, accountability and responsibility.
What I don’t know and couldn’t see in either organization is the ethical behavior of the executives at defense contractors. What holds them accountable? How much room is there for them to cheat tax payers?
"Nobody acts with the intention of doing away with accountability."
This sounds optimistic. I do believe that if banks have such armies of incompetent compliance teams, it is because the management believe their own accountability is diluted this way. The incompetents do not know what accountability is and believe that they are giving advice and guidance to those who are accountable (the ones who think they just shed their own accountability away).
Highly regulated industries end up acting as government bureaucracies do for the same reasons but I would not exclude intention on the part of some (a lot of) individuals.
Two Kling-themes that are in tension with the possibility of reaching the goal of improving accountability by means of clear job descriptions crafted specifically with that goal in mind.
Over 13 years ago you had a short post at Econlog, "The Job-Seekers Paradox" in which you made the point that many job-seekers want jobs that are as well-defined as possible, but the better defined a job is, the easier it is to automate (or outsource), so the less likely it is to continue to exist for long as stable employment for a local human. So, given that the capacity of automation technology keeps improving, by selection pressures, the labor market evolves to make the average job less and less well-defined.
The second Kling-theme - at least 15 years old from posts on Econlib - is to expound upon Garett Jones's observation that most workers don't make widgets anymore, instead they are producing organizational capital (or their work could be categorized as 'overhead'), and that it is inherently difficult to assess an individual employee's output in terms of the value added to the organization that could translate into marginal product that determines wages, and this gets increasingly difficult with ever-increasing complexity and specialization.
My point is that these trends, which are continuing to strengthen in significance because driven by technological developments, make it less likely that a sustainable local human job CAN be held accountable, as individual performance becomes harder to assess, and duties cannot sufficiently well-defined to provide the necessary basis for accountability-maximizing craftings of job descriptions.
I suspect that in addition to employees trying to evade accountability by adopting bureaucratic structures and procedures that diffuse responsibility, that this increased difficulty in holding workers accountability is also happening as a fully spontaneous consequence of the changing nature of the work itself, as a kind of emergent phenomenon that appears as the capabilities of automation technology pass a certain level.
One prediction is that with that capability looking to suddenly grow fast with AI, that the substitution will occur disproportionately precisely to the jobs easiest to hold accountable, and thus the jobs that most humans will do will rapidly become inherently resistant to any efforts to hold them accountable by defining standards of performance.
So, in the near future, every job will become increasingly like a government job. Have a nice day!
Bureaucracy leads to less output, or more inputs for the same output. Also delays. It also prevents innovation, new product discovery, etc.
Main reasons:
(1) chain of command approvals/committee approvals
(2) No awareness of applying rules to unique cases
(3) Turf wars
(4) Budget expansion and more employees yearning for direct reports
(5) Paperwork: minutes, memos, forms
(6) scheduling issues, people fighting to be kept in the loop
(7) Pettifogging: undue emphasis on petty details
I've a general bureaucracy theory which I bet someone smarter has already invented, and has been named after someone else.
1. Start a small company. One person handles all HR and office functions: tracking vacations, sick time, pay rates, new hires, old fires, stocking the break room. Probably has a real job too.
2. More people are hired. At some point, the HR person is working 50 hours a week, so the company hires an HR clerk. Now they together work 60 hours a week.
3. Wouldn't do to have the new full-time clerk only work half time, or loan him out to other departments, so the HR boss finds new work. Organize a Christmas party, river and trees outings for morale, monthly lunches and go kart afternoons.
4. And so on. Company eventually gets to 100 employees and 3 in HR.
5. Then trouble. Market changes, new competition, economic downturn. Lay off production and design employees, but HR? Not on your life! Employees have come to expect go kart afternoons, and too much cutback leads to industry rumors of not doing so well.
6. But markets force changes, if businesses won't make them voluntarily.
And 6 is where government goes off the rails. They are monopolies.
The US created a tea importation board in 1897 to verify "that the tea should be rejected if it was unfit" which bureaucrats of course expanded to include tea tasting. It was finally abolished in 2023, 27 years after its budget had been zeroed out and was no longer taste-testing tea.
https://reason.com/2024/03/17/after-a-century-the-federal-tea-board-is-finally-dead/
There's also the possibility of structuring government agencies as corporations, or farming out government operations to private bidders. There are problems with this approach, but are they worse than the problems associated with unaccountable agencies? Prior to the 20th century, the state often compensated public servants by facilitative payments and bounties. To cut wasteful spending and improve public services, perhaps we should consider partial reversions to the old system: https://www.amazon.com/Against-Profit-Motive-Revolution-Government/dp/0300194757
Arnold Kling writes about position descriptions today. He advocates using the PD for “aligning authority with accountability.” He observes that “In government, the tendency is to minimize the authority of any individual. Instead, a typical action is subject to a complex approval process. This makes it very difficult to assign accountability.”
However, a new central government political appointee might also need to be aware of the roles of the PD in the position classification system under chapter 53 of title 5, United States Code, (https://uscode.house.gov/browse/prelim@title5/part3/subpartD/chapter51&edition=prelim )
and in the performance appraisal system under chapter 43 of title 5, United States Code (https://uscode.house.gov/browse/prelim@title5/part3/subpartC/chapter43&edition=prelim ). Further, the new political appointee should be aware of the organizational alignment of authority and accountability supposedly provided under the strategic and performance plans originally required by the Governmnment Performance and Results Act (see 5 U.S.C. 306 and 31 U.S.C. 1115) with which individual performance appraisal plans are supposed to be in alignment. In fact, it would probably behoove anyone who is going to be supporting or helping the Department of Government Efficiency to become well acquainted with title 5, United States Code generally: success will likely turn on mastery of the bureaucratic procedures authorized there.
However, eager reformers should not be intimidated by the morass, but rather take heart in knowing that government efficiency is embedded in the statutory merit system principles: 5 U.S.C. 2301(b) provides in part:
“(4) All employees should maintain high standards of integrity, conduct, and concern for the public interest.
(5) The Federal work force should be used efficiently and effectively.
(6) Employees should be retained on the basis of the adequacy of their performance, inadequate performance should be corrected, and employees should be separated who cannot or will not improve their performance to meet required standards.”
(https://uscode.house.gov/view.xhtml?req=granuleid:USC-prelim-title5-section2301&num=0&edition=prelim )
And moreover, it is a prohibited personnel practice under 5 U.S.C. 2302 for any manager or supervisor, career or not, to take personnel actions to prevent an employee from disclosing waste:
“8) take or fail to take, or threaten to take or fail to take, a personnel action with respect to any employee or applicant for employment because of-
(A) any disclosure of information by an employee or applicant which the employee or applicant reasonably believes evidences-
(i) any violation of any law, rule, or regulation, or
(ii) gross mismanagement, a gross waste of funds, an abuse of authority, or a substantial and specific danger to public health or safety...”
(https://uscode.house.gov/view.xhtml?req=granuleid:USC-prelim-title5-section2302&num=0&edition=prelim )
The first thing to know about position descriptions in the central government is that they are the principal document upon which pay for the position will be determined. Therefore, the document has to make all sorts of outlandish claims about importance and difficulty in order to support the pay grade of the position’s supervisor. 5 U.S.C. 5104 (https://uscode.house.gov/view.xhtml?req=granuleid:USC-prelim-title5-section5104&num=0&edition=prelim ) sets out the general criteria for grading positions.
At the top of the General Schedule pay system under which most central government employees are paid, is the GS-15 grade. 5 U.S.C. 5104 states:
“(15) Grade GS–15 includes those classes of positions the duties of which are-
(A) to perform, under general administrative direction, with very wide latitude for the exercise of independent judgment, work of outstanding difficulty and responsibility along special technical, supervisory, or administrative lines which has demonstrated leadership and exceptional attainments;
(B) to serve as head of a major organization within a bureau involving work of comparable level;
(C) to plan and direct or to plan and execute specialized programs of marked difficulty, responsibility, and national significance, along professional, scientific, technical, administrative, fiscal, or other lines, requiring extended training and experience which has demonstrated leadership and unusual attainments in professional, scientific, or technical research, practice, or administration, or in administrative, fiscal, or other specialized activities; or
(D) to perform consulting or other professional, scientific, technical, administrative, fiscal, or other specialized work of equal importance, difficulty, and responsibility, and requiring comparable qualifications.”
As of March 2024, the central government’s Office of Personnel Management reports on its FEDSCOPE statistical system that of the central government’s 2,278,730 total employees, some 71,780 individuals, or about 3% of the total workforce, encumbered positions graded at the GS-15 level. The first lesson, then should obviously be that the federal pay and classification systems and everything about them are a farce. It is probably much worse on the military side where one can’t swing a dead cat without hitting a flag officer and the ratio of enlisted to officer is wholly absurd. Total central government payroll obligations for 2023 amounted to $686,716 million or over 10% of net federal outlays of $6,134,672 million for that year.
In the central government personnel system, the performance appraisal system is where authority and accountability have a putative alignment. Yet rarely will the individual performance appraisal plan align with the position description of the incumbent in terms of all the exceptional and outstanding duties assigned to the position. In fact, it all dissolves into a hopeless muddle. And for a large share of central government employees, the performance appraisal output measures are negotiated in union contracts. A bit of sunlight might disinfect this sordid state of affairs, and DOGE might well consider requiring agencies to post all of their union contracts alongside their budget documents on their agency websites.
Because position descriptions are not aligned with performance appraisal plans, government reformers have tried to impose top down accountability with things like the oft amended Government Performance and Results Act which list an agency’s work plan and the outcomes for which it intends to hold itself accountable. These documents too are risible but they are a good place to start in figuring out why statutory programs are being neglected in favor of nebulous feel good performance measures for which accountability is largely impossible.
Plus government arguably has extra failure modes. Like how do we limit authority while adding accountability? A strongman or Spoils system president has lots of accountability in principle, but little in practice. Process accountability, vs pure outcome accountability, is expected not only to foster reliability but to limit corruption and power grabs.
I guess it comes back to big public choice dilemmas like lack of competition (only one federal gov), complexity (fuzzy relationship between policy and outcome), voter incentives... A corrupt CEO ultimately faces shareholders. And i feel like the libertarian answer of limited government kind of begs the question.