How many direct reports should an American president have?

Arnold Kling writes:

One way to improve government operations would be through re-organization. I once wrote,

“the total number of executive entities is 157. I cannot think of any corporation in which the CEO has so many direct reports. This number ought to be fewer than ten.”

I proposed consolidation. Ideally, this would be done through legislation. However, if Congress balks, the President could informally choose to make some Cabinet officials and agency heads subordinate to others

Such informal hierarcies arise in any case, but can you imagine a private corporation that tried to run on such a basis?  NB: You can and should be horrified by this organizational detail, without adhering to the (false) view that “government should be run as a business.”  If you wish, take a large non-profit and ask how many people are direct reports to the CEO, or ask how such organizations would fare if the hierarchies of responsibility were never outlined explicitly.


'but can you imagine a private corporation that tried to run on such a basis?'

Considering the size and reach of the government's activities - think satellite fleets, for one example of an extremely complex and high tech endeavor where the government itself uses different 'executive entities' (since I have no interest in the AEI's capcha process, forgive the following if incorrect) such as NOAA, NASA, various parts of the Defense Department (such as NRO or Air Force or NSA or National Geospatial-Intelligence Agency or Missile Defense Agency), the CIA, and likely several other agencies (think Landsat), it is clear that making a comparison between the American government and a private corporation is what is unimaginable.

Unless a private corporation can be imagined as being in charge of the totality of America's nuclear deterrent, from aircraft carrier battle groups to submarines to bombers to MIRVed ICBMs. Which is only facet of what the American government does, of course. It would be interesting to see how many of those 'executive entities' are connected to the president's role of being the commander in chief, which is not some sort of equivalent to a CEO.

"since I have no interest in the AEI’s capcha process,"

Looks like a failure for a variant of the Turing test.

Or since I would need to explicitly to turn on both images and javascript and likely cookies (and possibly even use a profile that sends a HTTP referer), there is no reason to waste my time. This web site works just fine using state of art web technology ca. 1995 after all - much like the Post, excluding the HTTPS aspect.

A text based web, without any images or flash movies, is just fine for me, and why I continue to be such a devotee of Seamonkey and its ability to simply ignore decades worth of web development devoted to allowing advertisers to more effectively make money. Not to mention it is much, much faster, and never crashes - the security benefits of not using javascrtipt are just a bonus.



Don't worry, few people have any concern about their privacy online, much less leaving attack vectors open.

It's considered good emergency management practice to have no more than seven people directly reporting to the chief.

150 to the tumbrels! ;)

Actually it's good marketing practice to have the answer be "7" to any numerical question regarding wisdom. It's an answer that feels right to people.

It also feels "right" to starlings---and for good reason!

> “the total number of executive entities is 157. I cannot think of any corporation in which the CEO has so many direct reports. This number ought to be fewer than ten.”

That this is a good solution is not nearly as obvious as Kling makes it out to be. Or rather, it is a fine solution; what he does not mention is that the solution creates all sorts of new problems. For example, adding additional layers of hierarchy may solve the problem of information overload among executives. But it also means longer chains of communication between the top and the bottom of the organization, which create new possibilities for information distortion, etc. Herbert Simon pointed out the dilemma already back in 1946, in an article entitled "The proverbs of administration":

"Administrative efficiency is supposed to be enhanced by limiting the number of subordinates who report directly to any one administrator to a small number - say, six. This notion that the "span of control" should be narrow is confidently asserted as a third incontrovertible principle of administration. The usual commonsense arguments for restricting the span of control are familiar and need not be repeated here. What is not so generally recognized is that a contradictory proverb of administration can be stated which,
though it is not so familiar as the principle of span of control, can be supported by arguments of equal plausibility. The proverb in question is the following: Administrative efficiency is enhanced by keeping at a minimum the number of organizational levels through which a matter must pass before it is acted upon" (Simon, 1946:56).

This is an interesting discussion, thanks.

Maybe the conventional wisdom on hierarchy needs revision--but still, surely 157 direct reports is well beyond the optimal number.

It's not surely - maybe it is, maybe it isn't. The point is that the conventional wisdom on hierarchy doesn't apply, because public corporations are much smaller than the government.

There is no way that a single person can spend enough time and know what is going on enough to have meaningful oversight over 157 people. That was my point.

To the extent that the status quo is working, it is working because de facto hierarchies have been put in place.

'There is no way that a single person can spend enough time and know what is going on enough to have meaningful oversight over 157 people.'

And yet, bizarrely, your typical high school teacher, with 5 periods a day with 31 students per class, is in exactly this situation. No wonder the U.S. is in such a mess.

Come on. I meant meaningful managerial oversight. A student-teacher relationship is totally different.

Isn't the point of nominating cabinet members so you don't need to have "meaningful oversite"? You nominate someone you trust (presumably trust more then yourself since no one person can be an expert on every thing)

Government is special because, while it's very large, the very top is there for a relatively short period of time. A good part of why large organizations have to act in inefficient ways is that time and complacency creates an environment with no shared purpose.

A startup, in comparison, has a few years where it can underinvest in management. 50+ direct reports which self organize, working in a way that makes traditional structures appear unnecessary. Since the top of the government has such short tenures, they can copy this reduced oversight and be more effective than, your typical 30+ year old fortune 500.

The point is a head of state, like any other executive, has a finite capacity for effective management of direct reports. So if you want to address the problem, either limit the reports, reduce the size of government, or break up the country into more manageable chunks.

If you broke the country up, you'd still have national governments of similar scope (if not greater because less was farmed out). You'd just have less contention and less activity on the part of obstructive veto groups.

If you want to decentralize in this country, the simplest way to do that on short order is to limit inter-government transfers (leaving aside the insular dependencies and the Indian reservations for a moment) between the central government and state and local government to (1) small rental payments and fees consequent voluntary contracts (e.g. the feds renting office space or land), (2) small indemnities awarded by administrative tribunals or federal courts (say, when a federal regulation compels a capital expenditure by a local government, (3) disaster relief, (4) payments into state trust funds for long-haul Interstate maintenance, financed by tolls, (5) unemployment compensation, (6) Medicaid, and (7) an unrestricted subvention paid to state and territorial governments and distributed per formula (meant primarily to help impecunious states). Except for items 1, 2, and 3, you'd have no direct federal-local payments at all and local public policy would be set with scant reference to federal policy bar the usual harassment from the federal courts.

You could add payment-in-lieu of taxes on federal property to the foregoing list.

Good points. Doesn't it all hinge on A) which decisions actually need presidential input (in some ideal universe) and B) which decisions actually come before him?

This probably varies over time, place, technology, geopolitics, etc., but in theory there could be a bureaucratic culture that makes appropriate judgements about how high decisions float up the chain and leaders that trust such autonomy. We can dream...

I'm glad someone has made this point. It's not obvious that adding two or three more layers of formal hierarchy would help make things more efficient.

In practice, aren't a lot of these things delegated anyways? Where the formal structure doesn't exist, you get an informal structure instead, with cabinet secretaries needing to go through the Chief of Staff (or whomever) before they can talk to POTUS. Are these people direct reports in practice? And people can be "direct reports" to POTUS without actually directly reporting anything to him for months on end depending on what's going on at the time. You call into the room the people who are relevant at the time and otherwise let the bureaucracy do its thing. Regardless, having this many "direct reports" allows each president to choose how they want to structure their reporting chain in practice without requiring new org charts and the accompanying controversies and hurt feelings.

"Are these people direct reports in practice?"

No. Most have no regular access to the president. Even full Cabinet members may have difficulty getting access.

"69 independent agencies and government corporations, 69 boards, commissions, and committees, and 4 quasi-official agencies"

Some of these probably cannot get White House staff access.

I guess you could make an argument that by leaving the reality so color-by-numbers, you risk the potential of an executive who isn't very good at administering large organizations not being able to get something set up right. But I think the costs of one-size-fits-all outweigh the benefits here.

The government is more like a corporation with the president as CEO, but many executive functions reporting primarily to the board (Congress in the case of government). This doesn't really work in a normal corporation, but when Congress creates an independent agency with a separate budget (and out of the president's control), many of the concerns of ordinary corporate governance are avoided. And, in principle, Congress should be applying much more oversight than a traditional board. It doesn't mean this is the right arrangement, but you can't argue purely by analogy here.

Wikipedia states that the Catholic Church as over 200 cardinals. Seems to work for them.

The Catholic hierarchy is the government of an imaginary kingdom and doesn't govern anything. So any rule that "works" for it should probably be reversed before applying it to real governments.

Part of the problem arises from the unspoken assumption that "efficiency" is an equally important goal in both public and private sectors. That is a false assumption.

Efficiency is critical to a company; it's an ingrained principle of corporate strategy that the low-cost producer has an existential advantage.

But when it comes to statecraft, or military power, or tax collection, the efficacy of result is far, far more important. Who cares how "efficient" your government is if it gets you into wars and then loses them?

This means that rules of thumb like ratios of direct reports – which are built to optimize efficiency – are at best irrelevant, and at worst positively harmful.

Is 157 the right number? I don't know, but the right solution is to hire really, really good Departmental Secretaries, and "manage" them only at the policy level – not the metrics-obsessed microscopic obsessiveness that characterizes most private sector efficiency-driven models.

Francis Fukuyama's book Trust has a section on the former computer maker Wang Computers. Where the boss did not trust anyone and so had a similar number of people reporting to him. It worked while the founder was in charge. For his son not so much.

The British colonial government of India called this paperlogging. The idea was that any new Governor-General would come in with lots of new ideas but drowning him in red boxes would result in him giving up early.

Trump just needs to slash and burn. Why does Washington have a Department of Education? Get rid of it.

If we are going to start a purge of departments, why not start with totally pointless departments? What functions of the VA couldn't be handled by HHS or DOD? Most of Energy's functions could slot under DOD. You could merge Interior and the EPA.

The VA is essentially a socialist health care, housing and comoensation system for more than 20 million people. While you are correct that DHHS seems a fitting department for this, I submit that the VA is too large to be folded in with DHHS. In fact, DHHS would become infected with the ideals of entitlement. Whereas the VA has a "you break it, you bought it" rationale, DHHS is founded upon more general welfare arguments. The latter is more easily reformed. Americans largely agree we OWE veterans care even though we don't agree we owe the poor a handout.

There will be no helping people except on the part of those who can be motivated by guilt tripping.

Standardized means of assistance are basically the equivalent of Stalinist starvation in Ukraine.

Often agencies are not 'totally pointless' because they have distinct institutional cultures and aims. The Hoover Commission was criticized by some quondam officials for failing to recognize this. Someone defended the Farmers Home Administration thus: "Our mission was to make bad tenant farmers into good tenant farmers [by offering credit and extension services], not to deny them credit because they were bad farmers [which is what the Farm Credit Administration would have done]".

DOE is mostly a superstructure for the National Laboratories with a mess of other functions thrown in. It acquired a reputation for mismanagement within a few years of its creation and one former secretary said only someone 'probably certifiable' would want the job for more than a few years. It's begging to be busted up,

Interior is a service department primarily concerned with resource management. EPA's core function is health and safety regulation. It does a mess of other things too, some of which could be discontinued (grant programs) or transferred to interior (environmental public works). However, they're distinct in that one is predominantly regulatory and one has a regulatory aspect.

=== "slash and burn"

Exactly. The monstrous US Federal Government has grown far too large to be managed by any rational organizational system. It is quite literally out of control.

US Presidents are political figureheads. They have no real management awareness nor control over anything, other than (perhaps) their top layer White House staff.

Kling has no clue how the people in Federal departments actually operate day to day.

True dat.

But maybe that is Klings ultimate point. The president CAN"T control the government. The "direct report" argument is somewhat trivial to the nature of the problem, as you state. Even if the president reorganized his cabinet to have division and subordinate commanders, he would still lack control over day to day operations because of agency independence.

The political question is whether we as a nation WANT agency independence from the president. If so, to whom are they accountable? Congress has committees and subcommittees specifically set up for oversight and accountability, but it is overwhelmed with its other duties and political maneuvering. The agencies are hence a de facto fourth branch of government, unelected and unaccountable.

A larger budget for congressional oversight functions?

With more rigorous mandates? Or less rigorous and more free mandates?

Agencies that don't have any law enforcement, regulatory, foreign affairs or military function don't necessarily need direct supervision from the President. If they are wasting money or failing to achieve their objectives, Congress can step in.

Ivanka does Education now, so you can take that off the list.

Lots of candidates for excision:

1. The discretionary grant facility of just about every agency there is. That will mean liquidating agencies that do little else (the Corporation for Public Broadcasting, the National Endowment for the Arts, the National Institute on Food and Agriculture, the Small Business Administration, the Economic Development Agency, the Minority Business Development Agency, &c.).

2. The Department of Housing and Urban Development. Send the lead-paint regulators to the EPA, and liquidate the department and its programs.

3. The Department of Education. Send the survey research function to the Labor Department, send a small selection of regulatory functions to a stand-alone commission or to the Federal Trade Commission, and send residual student aid programs to a resolution agency to wind down. Liquidate the rest of the department.

4. HHS: retain the Center for Medicare and Medicaid Services, the National Institutes of Health, the Indian Health Service, the Commissioned Corps of the Public Health Service, the Centers for Disease Control, the Food and Drug Administration, and the Office of Refugee Resettlement. Shut down everything else. Be sure to gut the grant facility of NIH and CDC to zero.

5. The USDA: retain the regulatory agencies, the in-house research service, the statistical collection services, the pest control service, the soil conservation service, the Forest Service, and perhaps the extension service. Shut down everything else (notably the SNAP program which accounts for 2/3 of the budget, the farm subsidies, &c).

6. The Transportation Department: most of the budget consists of subventions to state and local government (with some side items like subsidies to motor carriers and concessionary tolls to general aviation). Set up toll booths to finance long-haul Interstate maintenance and distribute the proceeds to the states on a per-acres-of-macadam basis. Distribute signage for the Interstates and U.S. Routes. Otherwise, abolish the grant programs.

7. The Department of Energy. It's always been an administrative nightmare. Again, end the grants and appoint a blue-ribbion commission on the future of the in-house and contract research conducted by the National Laboratories. Sell the loan portfolio, transfer the armory to the Defense Department, transfer the site-clean-up programs to the EPA or the Interior Depeartment, transfer the statistical agencies to the Commerce Department, transfer the intelligence office to the military or Homeland Security, transfer the regulatory services to a stand-alone agency, and transfer the electric authorities to a stand-alone public corporation. What's left would be the National Labs, which could be run under the auspices of a stand-alone agency.

8. The VA: transfer the cemeteries to the Interior Department, transfer large blocs of physical plant and staff to newly-incorporated philanthropies governed by elected stakeholder boards, and institute insurance and voucher programs in lieu of most (if not necessarily all) in-house care offered by the Veterans Health Administration. Also, simplify miscellaneous benefit programs, offering primarily cash doles, scholarships for education and training, medical care, long-term care, and a grave. Cut out ancillaries like the residential mortgage program.

9. Miscellaneous targets for the auction bloc or closure: the Postal Service (bar the Postal Inspector), the Farm Credit System, the ExIm Bank, the Corporation for National and Community Service, the shizzier foreign aid agencies, &c.

10. The Department of Justice: send the investigatory services to Homeland Security and append the federal prosecutors and some odds-and-ends to the judicial branch.

11. The Department of Labor: retain the Job Corps but shut down the other training programs.

12. Civil rights agencies: close them all down.

I still like Lew Rockwell's 30-Day Plan. This is a good list too.

The vast majority of NIH's work and impact is through grants. These are the basis for the majority of drug discoveries that make it to market.

These are the basis for the majority of drug discoveries that make it to market.

You've got the talking points, I see.

Research is a public good and grants are a way to outsource research work. Cutting funding for research entirely is dumb and the federal government does not have any clear expertise in hiring and managing top Ph.D. and M.D. researchers.

Research is a public good

It is no such thing.

All these magic public goods, with no diminishing marginal return!

You've put some thought into this! I agree with almost none of this, but thanks for a thought-provoking post. Let me make an argument in favor of the grants: They're grease on the wheels of government. Congressman John Smith (R-Podunk) may hate big government, but he loves federal largess coming into town. I imagine you see this as a negative, but I think so much of the federal government happens so far removed from the average citizen that they can't evaluate it at all: I cannot possibly tell you whether the F-35 is really a good investment for our country, but I'm all in favor of the EPA continuing to fund the toxic cleanup at our port. I think people need to be able to see a bit of federal spending with their own eyes and feel it beneath their feet. I could make a similar argument in favor of earmarks - I think our political system actually operates less efficiently with less payola. Anyway, kudos on the ideas even if I don't agree.

You misread Fukuyama's example. Wang was a case of TOO MUCH trusting, not too little.

His point was that the Chinese model of leadership was family-based, hence An Wang promoted his son – an unfortunate move, since Fred Wang turned out to be a disastrous leader.

Your point about "Trump needs to slash and burn" is a complete non sequitur. Fukuyama and Wang have nothing to do with your claim. Which I think is also wrong, as per my comment above. We don't need an efficient State Department, we need an effective one.

Who do you think An Wang trusted too much? Not his senior executives as he tried to gather all the power in his own hands.

And yes, I noticed, and I said, his son could not manage it. Which is the point. A founder, a genius, can have that many people reporting to him. But a more ordinary person - the sort of person likely to win an election - needs fewer people reporting to him and more executives he can trust.

So it follows perfectly well from the Wang case. Efficiency and effectiveness go together. Certainly having the Democrats entrenched in the State Department running their own policy is not likely to work

Even slashing and burning requires management. Congress is not going to stop allocating money just because the president says so.

Kling is looking at org charts for his data.

Only an academic would look at an org chart, rather than looking at how people actually report or act in the real world.

Knowing how the real world works would lead one to believe a contrary fact: 1) Being a direct report to the President in most cases means you have full authority, or you simply report to some peon in the EOB; 2) Your place on an org chart is reward or trinket offered to you to attract you to the job. If you were really interested in reporting, you would track the email and correspondence traffic and use that network data to see the real persons who actually direct report to the President.

I have only one direct report.

I have a direct report to my wife

Unless your single direct report is again your wife, you're in trouble, man...

Best trouble I ever had.

Your wife is your superior?

Good grief. Man up, dude. Take off the skirt and give it back to her.

We each have our own domain.

Score another one for Trump: now that he's President, things that have long been obvious are suddenly occurring to people who make it a priority to ignore the obvious.

Right. Good to see that you're coming around to acknowledgement of Russian influence in the White House.

What you academics have discovered is this: actually Congress runs the government in most respects. And each of those 157 reports in effect to a co-CEO (committee chairs in Senate and House).
But there's no point in posting this; you won't read it dear authors and if you read it you won't really understand it. Sigh.
Meanwhile, the President has very few direct reports in any meaningful sense. Well, the current incumbent is a mystery but the previous ones, all of whom I have known since 1993, played the role in a reasonably similar manner and had very few direct reports. I could go on but there's no point because you won't pay attention and if you did, you'd choose quirky inaccuracy over understanding.

Yes, a lot of the reports to the President mentioned by Arnold Kling are independent agencies set up by Congress like the National Transportation Safety Board. The NTSB gets on with investigating air crashes etc and the President has little to do with its day to day operations.

The Board is a perfect example of agency independence. That's why it's called a Board and not a department. It is set up to be mostly autonomous. The president doesn't even have the authority to fire the board members without congressional approval.

The NTSB is a tiny agency of professionals with intense skill sets. Its best left alone.

If you want to see pathology in the operations of the federal executive, dust off the old academic article titled, "Jamie Whitten, Permanent Secretary of Agriculture).

It's a lot easier to buy the government when there are only two parties.

This view is correct. While the president appoints the agency heads, the agencies are ultimately answerable and under control of Congress.

I am sure this is correct. They don't really report to the president, but to his staff, and only select numbers of his staff report directly to him, so this filtering process already exists.

Congress doesn't run the government. It influences policy through control of the appropriations process or authorizations. It's still administration appointees they're dealing with. How much do congressional committess actually 'run' John Koskinen's operation?

The goal of reorganizing government (or any organization) should be a better government (or organization). Reducing the size of or eliminating government agencies at random doesn't make a better government, just a smaller government. But only temporarily. Eliminating the department of defense won't permanently reduce the size of government because it will soon be apparent there's a need for the department of defense. And that's true for many government agencies, including those I may not appreciate. Attitudes about government often reveal a paradox. For example, the west coast and the east coast contribute more for government than they receive in return from government, but residents in those areas generally support government. Florida receives more from government than contributes for government, but residents in Florida generally oppose government. Mr. Kling's approach to reorganizing government is to cut the numbers, which numbers he doesn't say, just cut the numbers. When a surgeon enters the operating room she does so with a specific surgery in mind, the surgery needed by that particular patient. Sure, she could randomly choose a surgery for the patient, but that wouldn't improve the patient's health. Government is like that. Reorganizing government requires a good surgeon.

Agreed. It's one thing to say we could gain some efficiencies from a reorg (probably true). It's quite another to blindly start hacking off limbs.

There are masses of barnacles to be scraped off, the pet projects of members of Congress long departed. (Claiborne Pell saddled us with the immortal NEA).

Did you read Kling's post? He provided very specific examples of consolidation that eliminate redundancy and improve coordination. Not random at all. Why did you devote so many words battling a strawman?

How does Kling know if consolidating separate agencies will eliminate redundancy and improve coordination? His first idea for consolidation, incorporating NSA and CIA into Defense, may well be a terrible idea, since the purposes (and jurisdictions) of those two agencies are very different. A pithy list of eight government agencies to replace all of them probably appeals to Kling's belief that government is on the whole a waste. And eight is less than ten.

Is this supposed to be Straussian? Are we supposed to be thinking about Kushner's current list of responsibilities?

(Advising Trump, reinventing government, shadow diplomat, brokering mideast peace, solving the opiate crisis.)

Or is it a bit of escapism, a dream of an alternative America where virtue and pragmatism reign?

The fatal mistake in the analysis is that these are not typical corporate entities and the direct reports are not actually running day to day operations.

Agencies are headed by political appointees who certainly have a lot of influence, but operations are conducted by civil servants who are mostly autonomous.

Presidents have historically had a difficult time getting the bureaucracy on board. Each agency varies in its level of independence from the president based on its characteristics.

Two counterexamples are Obama and Clinton who used agencies to great effect toward their agendas because the interests of the career servants were aligned. Despite Clinton's claim to have "reinvented" government, the only agency whose budget he cut was Defense.

These direct reports also don't meet daily, weekly, or monthly with the president. They are more akin to colonial governorates or viceroyalties. The king pays no attention to them as long as they are not acting contrary to his wishes.

I don't see the "fatal mistake". Most of the points you make seem to confirm the analysis's conclusions about the status quo being dysfunctional.

The current status quo, certainly. But go back to Obama or Bush, how would you measure, using any quantitative method, that dysfunction?

Or is it all reasoning from personal philosophy, and thus "non-transferable?"

So a politicized bureaucracy is a good thing in your mind?

I did not say that at all. I asked how we measure "dysfunction" outside our sky castles.

I mean it's fine for a man to wake up, and while still in his bathrobe, and with his first cup of coffee, to shout "they're all doing it wrong!"

We've all been there .. but how do we prove it?

Measuring dysfunction is really hard. Never mind performance; we won't even be able to agree on what the goals are. And even if we can agree that things are not functioning, is it because of a breakdown at the executive level, or something else?

We tend to ignore this. Yet competence makes a huge difference and it might be considered low-hanging fruit (compared to intractable ideological disagreements).

I couldn't access the original AEI article so I can't tell what, exactly, is counted among these 157 direct reports but I suspect some of these people report to the Chief of Staff or Vice President in practice. It also probably includes people such as the heads of the Smithsonian or the Peace Corps, people who are basically CEOs of independent agencies who can be removed and replaced if something goes wrong. As pointed out above, Congress also plays an important role in supervising and checking agencies.

Who would have thought the day would come, when Arnold Kling and Dick Cheney would agree on something.

This reminds me of a parallel question that would apply to the corporate world:

How many divisions should GE have that report to the CEO, or Tyco?

Do we limit the size of a corporation or the scope of its activities based on the number of direct reports or divisions?

Bust 'em up because there are too many divisions and direct reports? Or, if you follow the logic of the post, maybe each division should have its own board of directors, just like Congressional committees oversee various parts of the executive branch.

Wouldn't that be wonderful.

Or not.

While it is true that adding layers of bureaucracy may offset any gains to be had from reducing the number of direct-reports to the President, consolidation as Kling recommends, does at least have the potential for other benefits. For one, just as mergers and acquisitions have relentlessly pruned white collar business jobs and thereby contributed to efficiency and productivity, so too could departmental mergers allow for the elimination of purely administrative expenses that would be made redundant by reorganization. Kling's proposals should be read taking into account President Trump's executive order requiring a comprehensive plan for reorganizing the executive branch. The comprehensive plan is to consider:

"(i) whether some or all of the functions of an agency, a component, or a program are appropriate for the Federal Government or would be better left to State or local governments or to the private sector through free enterprise;

(ii) whether some or all of the functions of an agency, a component, or a program are redundant, including with those of another agency, component, or program;

(iii) whether certain administrative capabilities necessary for operating an agency, a component, or a program are redundant with those of another agency, component, or program;

(iv) whether the costs of continuing to operate an agency, a component, or a program are justified by the public benefits it provides; and

(v) the costs of shutting down or merging agencies, components, or programs, including the costs of addressing the equities of affected agency staff."

as well as "other relevant factors."

The Executive Order provides for a public comment period. I hope that Kling will submit his proposal. Effective span of control should be considered as a relevant factor. Research into administrative overhead in executive entities has been neglected unsurprisingly, given most federal entities shell out money to academics one way or another. However, one need not believe in the virtue of smaller government to recognize that resources could be freed up for mission-related purposes by eliminating administrative overhead through agency mergers. Why unnecessarily pay for two sets of chief human capital officers, chief IT officers, public relations, congressional affairs, chief financial officers, general counsels, inspectors general, diversity and inclusion councils etc etc.? Even within a department, the number of direct reports is well beyond optimal. Look at the poor Deputy Secretary of Treasury with at least 18, for example: And many of the subdepartmental entities replicate the entire administrative panoply within themselves. President Trump has some low-hanging fruit here, ripe for the picking.

1. Assemble the extant corpus of federal agencies and authorities into about two-dozen departments governed by a secretary serving at the President's pleasure and a half-dozen trusteeships governed by boards whose chairman might serve at the president's pleasure but whose other members would serve fixed terms. The trusteeships might include the Federal Reserve and Mint, the Postal Service, the printing and broadcasting plant; the federal museum, libraries, and archives; the hydroelectric authorities; and miscellaneous enterprises.

2. Have oversight commissions (with members serving fixed terms) to departments concerned with tax collection, law enforcement or intelligence gathering. Have commissions serving fixed terms appended to departments whose function includes regulation, and have these commission be the ultimate authority in rule-making and adjudication. So, you might have a half-dozen regulatory commissions rather than the industry-specific antheap you have now (say, one for the financial sector, one for land and resource use, one for commerce NOS, one for labor relations, one for health and safety, and one for virtual property rights like copyright or broadcasting spectrum).

3. Provide for the president to appoint up to a half-dozen superintendents. Each would be assigned a portfolio of departments (or trusteeships) to supervise. The VP could be employed in this capacity.

4. Have the President's direct reports consist of the White House chief of staff, the superintendents, a discretionary selection of his cabinet secretaries, and the Chairman of the Joint Chiefs of Staff. A single-digit set ought to do it.

"appended to departments"

The different department heads should have to engage in mortal combat, with the winner taking over the loser's department. This will set a good tone for the underlings when the reorganization begins.

.. but with Jeopardy.

It worked for the Ottomans.

If the federal government would stop interfering in matters where the Constitution gives it no right to interfere, it could cut armies and armies of "reports".

What does he need? (i) Armed Forces. (ii) Securitate. (iii) Law. (iv) Liaison with Congress. (v) Foreign Affairs. (vi) Treasury. (vii) Liaison with States. (viii) Internal affairs and border protection; coastguard. (ix) Colonial affairs (but don't call them "colonies" for God's sake). (x) the one I overlooked.

There: ten's enough.

1. There are no colonies. The only dependencies we ever had which have seen demographically important settlement from the mainland were Alaska and Hawaii. They're fully integrated and the aboriginal population therein is a modest fraction of the total. As for the remaining insular dependencies, they have their own governments and federal functions peculiar to them are too spare to merit more than a subdepartmental agency devoted to said purpose. The Indian reservations require more intrusive federal administration and have a larger population than the sum of insular dependencies bar Puerto Rico, but the three agencies dedicated to them have a budget in sum of less than $8 bn. (A contextually similar sum in the UK would be < 1 bn pounds).

2. The coast guard, border guard and point-of-entry inspectorates are notable for not being concerned with 'internal affairs;.

3. We have no central liaison with the states at all, nor is their any set of functions which would be enhanced by it.

4. Intramural functions require not only a treasury, but a personnel bureau, a plant and equipment bureau, and a records and IT bureau. Also, treasury functions which are intramural are distinct from revenue collection, which is antagonistic to the public and requires an intrusive inspectorate.

5. Foreign affairs has three components with distinct skill sets: the diplomatic and consular service, the civilian intelligence services, and the overseas development and relief apparat.

6. The congressional liaison officer is a member of the White House staff with a few aides. The position's been around since 1974 if not earlier. It doesn't merit a distinct department.

7. Nicolae Ceausescu never lived here and the federal police resemble his in no particular. The federal police are a collection of specialty services scattered across several departments as we speak. The only bureau with a non-specific institutional mission is the FBI, which has a budget of about $8 bn. (again, in significance < 1 bn pounds in a British context)

8. The military also incorporates most of the intelligence apparat as measured by budget and personnel.

The federal government has a large inventory of land, as well as proprietorship of the rivers, coast, game, and the air. It also operates the air traffic control system. There are 40 long haul limited access highways which require some sort of common government. You also require a civil defense apparat, You'll require a welfare apparat for clientele poorly integrated into states as political communities or possessing a special status not peculiar to any given state (e.g. veterans, military families, people in itinerant occupations, reservation Indians, refugees). Quite apart from that, you'll need several regulatory inspectorates to set standards for genuinely cross-border trade, a central bank, a mint and printing and engraving plant. Also, people move around, so regulating and providing for old-age pensions is a federal function as well.

Oversight is the job of Congress, not the president.

In By The People, Murray writes that the average Cabinet Department has 22 layers of management, versus 6 in the large corporate sector. US corporations undertook an aggressive flattening program in the 90s. Washington did not.

Corporations have flattened by spinning off and outsourcing much of their operations. For instance, when you stay at a huge multinational chain hotel, the property is often owned by an independent firm and the people who clean the rooms and work at the front desk work for yet another company that specializes solely in hotel operations (and their paychecks probably come from yet another company like ADP that specializes in payroll services). Government can do this and does in many situations: the DoD relies heavily on private contractors, the IRS often subcontracts collections work to private law firms, etc.

However, sometimes low-level government employees have lots of power -- think of customs officers or low-level FBI agents scattered throughout the country -- and it is simply necessary that they be part of a strict chain of command. Federalism means that a lot of government functions have already been flattened.

"layers" is probably not the right word.

Re fixes:

- have the government do less
- have it do more outsourcing (basically everything except the military)

Comments for this post are closed