I have some attraction to this model. The president would essentially become the appointer-in-chief. But I suspect that it would not actually get us all that far. The reason is that, from a libertarian perspective, it does not address the incentives for continued accretion of power in regulatory agencies. I think a better approach would be to address three problems, and then implement at least the chief auditor concept and a reorganization of regulatory domains.
The first problem is that regulatory agencies write their own laws by way of rule-making. If you believe that the regulators are more competent and knowledgeable than the Congress is, this should produce better laws. But we don't need the agencies to have legislative power for them to contribute that competence. For existing agencies, we could tell them to propose a set of laws that codify their rules and removes their power to set rules. For new agencies, we could let them write rules for some period of time, at which point their legislative powers sunset and they have to also propose codification of their rules. This avoids the ratchet effect of new rules being added when an agency gets new, more driven leadership. Perhaps both the regulatory agency and the president submit reports with codification recommendations. In order to make sure that something passes, we'd also need some changes inside of Congress's processes to make sure that Congress acted on the codification. We'd need to create an incentive for someone to propose a bill, privilege the consideration of those bills, and eliminate the filibuster for the passage of them. Perhaps we do a little crossing of the branches to allow the president to author a bill, and that bill that is privileged in that any member can call for it to be discharged from committee after 60 days, and the filibuster does not apply to it; or maybe those rules come into play only if a Congress does not itself pass something in its first session of a two-year Congress (assuming that the reports from the regulatory agency would be due within the first session).
The second problem is on the other end of the spectrum. There's no reason to believe that federal judges are incapable of understanding and properly interpreting the substance of today's regulations. Anyone who has ever been before a federal judge should know that they are some of the most fearsomely intelligent people you will ever run across. If a regulation is too complicated for them to understand (in order to apply), it is too complicated for anyone to understand (in order to comply with), and therefore should not be written. Meanwhile, agency-appointed administrative law judges have gotten a reputation for being shills. I think that moving form ALJs to federal judges is far more important than anything else in reforming administrative law. If a law passed requiring the agencies to submit a codification of their rules to Congress, that same law could require the agencies to include in their codification whatever jurisdictional provisions are needed to move from administrative law judges to federal courts.
The third problem is political accountability. The agencies tend to be led by commissions with fixed terms. I think we would be better off folding the agencies -- stripped of their legislative and judicial functions -- into the executive branch, with single directors appointed by and serving at the will of the president. That way, you can at least vote against the president when things go wrong, and that president can exercise removal power over the executive branch. If you couple that with at least the chief-auditor concept, you would have someone whose job it is to point out flaws and a politician who could act on that advice.
I'd like to add that I could see a role for a small number of well-regarded independent commissions to act as oversight. These would have no power other than the power to subpoena government officials and contractors (in those capacities) and to issue reports. We have a welter of such commissions currently. I'd suggest that they not be organized to map to the underlying agencies, but instead to address specific concerns. If we are worried about government policies promoting racial discrimination, have a commission that exercises oversight on the basis of discrimination. We could have one for corruption, environmental impact, and any other themes needed. But they would have to be limited in number (something like six) or they just become another voice. I'd give each member a single, nine-year term. The member whose term is about to expire would be chair. This probably ends up being much like the chief auditor approach, but having a plethora of potentially overlapping and competing auditors isn't a bad thing when government is its target.
I think the case that there needs to be a regulatory state is much stronger than the case that it needs to be dominated so much by federal agencies. Yes, there are genuinely interstate problems and efficiencies to be had from consistency of regulation across states; but is that really enough to explain the current proportion of federal vs state regulation? This matters in part because the more power is devolved to the states, the more impactful it becomes for some states to experiment with proposals like yours, which may be an easier way of getting things done than convincing federal politicians to support this sort of wonkery.
I note also that at least superficially, your proposed COO is similar to a city administrator position in a municipal government-- and some city governments have auditor offices as well, though perhaps with not such an aggressive mandate as you propose. So it's worth studying municipal experiences for lessons on how well this might work at a state or federal level.
I think this is a really good line of thinking, but I don't see how the COO/Auditor concepts would work without a lot of legal re-arrangements. And, if we were going to go down the path of making major legal re-arrangements, we could probably do better than installing a COO and CA. Not that I'd be opposed to it, but...
... I'd argue that most reform efforts should start with funding reform. Improve the budgeting process, which is manifestly broken. My utopian ideal would be for the government to spin off most of its "administrative state" functions to potentially competing non-profits. Taxpayers, after funding the basic government services, would be free to allocate their remaining tax dollars as they saw fit to he former government organizations, as well as any private venture that met basic qualifications to compete. If necessary, category of service targets could be put in place (e.g. at lest 25% of your tax dollars should be assigned to one or more of the 25 "Educational Service" governmental organizations. Imperfectly, it would create pressures similar to a true market for government services.
A beautifully written essay, but it is somewhat naive in that it posits management by disinterested agents committed to values of liberty and order. Instead, government has simply become the vehicle whereby the Hobbesian war of all against all is carried out. Regulation becomes its means as regulators become agents of predatory interests.
Is there evidence that more competent government leads to less burdensome regulation?
Let's take track and trace. Those governments that have been best at track and trace have had the most burdensome COVID interventions. The most lockdowns. The harshest impositions.
I'm not just referring to China here. Would any of us want the situation they have in Australia?
Even if we could have a government that was good at what it did, that wouldn't mean that "what they did" would be good. We might be better off with a government so incompetent that nobody wanted it to do anything. That seems to have been the case with COVID. Let it rip was by far the optimal solution with no close second.
I don't think competence is the major problem with government. It's that governments don't have credible signals of just what they should be doing to make peoples lives better. Capitalists have price signals. Governments have elections. One is a lot better than the other. It's not clear that taking away elections would improve the governments ability to figure out what it should be doing. Unelected Public Health Officials for instance have had way worse instincts than politicians.
Yes, Arnold, you can bet that all framers of past liberal constitutions never anticipated the world we live today. If you were familiar with the history of liberal constitutions in Spain, Argentina, and Chile, you would fear greatly how political competition for grabbing and holding power is being played today. Anyway, I think today's U.S. political competition is much worst than in each of those three countries for the simple reason that you have never faced the high degree of corruption of all liberal values and institutions I see in your country today. Too many have remained silent for too long in your country -- the only one with a good, long record of liberal values and institutions (yes, the poor record of those values and institutions in the last 200 years is a common excuse some intellectuals use to explain what is going on in Spain, Argentina, and Chile).
Regarding your proposal to reform just the bureaucracy, let me remember you that to reform radically the bureaucracy requires first a substantial reform of the Constitution -- but, as I used to discuss with Gordon Tullock in early 1980, reforming the Constitution at best is a necessary condition never a sufficient one.
This essay from November notes real problems; your ideas would likely reduce such problems. So I like them. I was really glad to see you note a couple of places the need for some term limitation, like 6 years. But don't believe they're realistic in the near term.
The Deep State "Fourth" branch needs term limits throughout. "Public Service" should be very limited for all. I'd propose that after 10 years, no more raises nor promotions (tho same-pay relocations OK), and conversion to a 1 year contract. Which can be renewed at most 1 time.
The "Civil Service" needs big reform - it ain't gonna happen.
Small, but very significant reform seems possible - like bureaucrat term limits. The single key metric for gov't agencies should be turnover - it should be 1% MORE than non-gov't wage earners (not small business owners). General IRS based turnover, reporting primary income from a different source this year than from the year before, should be a key number. Like 12% (or 10% or 15% or whatever it is).
Moving most gov't employees out of DC would also be generally good, and seems more politically possible.
I strongly fear that your ideas will be a case of "The Perfect being an Enemy of the Good".
I would support politicians supporting your reforms - would you support other politicians supporting other reforms that might make the situation less bad (/better ?) tho not as much as you propose here?
"Pick any problem the government is supposed to solve, and you will find the same problem: how do you measure success, and how do you identify how well the agencies should be doing?" is I think the key question. Governments lack the competition that markets provide, and so lack the mechanism for generating knowledge of how well the organization could be doing. Without being able to point at another org and say "Look, they manage to do X. Why can't you?" there is nothing to drive estimates of whether your division is doing well or not, other than just the vague feelings that drive our politics now. The COO and CA proposal does not contain a way to improve the status quo.
You say the administrative state is necessary to provide order; I say it hasn't been shown that the administrative state is even capable of providing order, much less necessary.
I also think that your proposal might fly if we were move the federal regulatory systems down to the state level, as at least there is a little more competition and comparison there. I don't think that solves the fundamental problem, however, and certainly it isn't the way current political currents flow.
I would like to point out something that is missing - you start with twin purposes for government in your first sentences - order and the promotion of virtue - and then proceed to only discuss the requirements of order and say nothing of virtue. Could you consider addressing it?
In order to create real accountability, your CA would likely need the power to fire individuals and close entire agencies. In fact, it might be best if the CA organization was required to terminate whoever it thought were the worst performing 5% of staff each year so that they can't dodge their responsibility for identifying them.
Idea in progress? Federal gov does have a GAO and aren't the Cabinet Secretaries supposed to be COOs? The questions seems to be not about creating new positions, but...(1) increasing the independence and effectiveness of the GAO and (2) making the huge Federal bureaucracies more transparent to Cabinet Secretary management. Problem: no one in the bureaucracies, the entities subject to regulation, nor Congress want either (1) or (2).
Arnold, the centerpiece of your proposal is purposefully putting successful businesspeople into high positions of government. In the recent past, what business figures who assumed roles in the executive branch have you most admired?
I have some attraction to this model. The president would essentially become the appointer-in-chief. But I suspect that it would not actually get us all that far. The reason is that, from a libertarian perspective, it does not address the incentives for continued accretion of power in regulatory agencies. I think a better approach would be to address three problems, and then implement at least the chief auditor concept and a reorganization of regulatory domains.
The first problem is that regulatory agencies write their own laws by way of rule-making. If you believe that the regulators are more competent and knowledgeable than the Congress is, this should produce better laws. But we don't need the agencies to have legislative power for them to contribute that competence. For existing agencies, we could tell them to propose a set of laws that codify their rules and removes their power to set rules. For new agencies, we could let them write rules for some period of time, at which point their legislative powers sunset and they have to also propose codification of their rules. This avoids the ratchet effect of new rules being added when an agency gets new, more driven leadership. Perhaps both the regulatory agency and the president submit reports with codification recommendations. In order to make sure that something passes, we'd also need some changes inside of Congress's processes to make sure that Congress acted on the codification. We'd need to create an incentive for someone to propose a bill, privilege the consideration of those bills, and eliminate the filibuster for the passage of them. Perhaps we do a little crossing of the branches to allow the president to author a bill, and that bill that is privileged in that any member can call for it to be discharged from committee after 60 days, and the filibuster does not apply to it; or maybe those rules come into play only if a Congress does not itself pass something in its first session of a two-year Congress (assuming that the reports from the regulatory agency would be due within the first session).
The second problem is on the other end of the spectrum. There's no reason to believe that federal judges are incapable of understanding and properly interpreting the substance of today's regulations. Anyone who has ever been before a federal judge should know that they are some of the most fearsomely intelligent people you will ever run across. If a regulation is too complicated for them to understand (in order to apply), it is too complicated for anyone to understand (in order to comply with), and therefore should not be written. Meanwhile, agency-appointed administrative law judges have gotten a reputation for being shills. I think that moving form ALJs to federal judges is far more important than anything else in reforming administrative law. If a law passed requiring the agencies to submit a codification of their rules to Congress, that same law could require the agencies to include in their codification whatever jurisdictional provisions are needed to move from administrative law judges to federal courts.
The third problem is political accountability. The agencies tend to be led by commissions with fixed terms. I think we would be better off folding the agencies -- stripped of their legislative and judicial functions -- into the executive branch, with single directors appointed by and serving at the will of the president. That way, you can at least vote against the president when things go wrong, and that president can exercise removal power over the executive branch. If you couple that with at least the chief-auditor concept, you would have someone whose job it is to point out flaws and a politician who could act on that advice.
I'd like to add that I could see a role for a small number of well-regarded independent commissions to act as oversight. These would have no power other than the power to subpoena government officials and contractors (in those capacities) and to issue reports. We have a welter of such commissions currently. I'd suggest that they not be organized to map to the underlying agencies, but instead to address specific concerns. If we are worried about government policies promoting racial discrimination, have a commission that exercises oversight on the basis of discrimination. We could have one for corruption, environmental impact, and any other themes needed. But they would have to be limited in number (something like six) or they just become another voice. I'd give each member a single, nine-year term. The member whose term is about to expire would be chair. This probably ends up being much like the chief auditor approach, but having a plethora of potentially overlapping and competing auditors isn't a bad thing when government is its target.
Max
I think the case that there needs to be a regulatory state is much stronger than the case that it needs to be dominated so much by federal agencies. Yes, there are genuinely interstate problems and efficiencies to be had from consistency of regulation across states; but is that really enough to explain the current proportion of federal vs state regulation? This matters in part because the more power is devolved to the states, the more impactful it becomes for some states to experiment with proposals like yours, which may be an easier way of getting things done than convincing federal politicians to support this sort of wonkery.
I note also that at least superficially, your proposed COO is similar to a city administrator position in a municipal government-- and some city governments have auditor offices as well, though perhaps with not such an aggressive mandate as you propose. So it's worth studying municipal experiences for lessons on how well this might work at a state or federal level.
I think this is a really good line of thinking, but I don't see how the COO/Auditor concepts would work without a lot of legal re-arrangements. And, if we were going to go down the path of making major legal re-arrangements, we could probably do better than installing a COO and CA. Not that I'd be opposed to it, but...
... I'd argue that most reform efforts should start with funding reform. Improve the budgeting process, which is manifestly broken. My utopian ideal would be for the government to spin off most of its "administrative state" functions to potentially competing non-profits. Taxpayers, after funding the basic government services, would be free to allocate their remaining tax dollars as they saw fit to he former government organizations, as well as any private venture that met basic qualifications to compete. If necessary, category of service targets could be put in place (e.g. at lest 25% of your tax dollars should be assigned to one or more of the 25 "Educational Service" governmental organizations. Imperfectly, it would create pressures similar to a true market for government services.
A beautifully written essay, but it is somewhat naive in that it posits management by disinterested agents committed to values of liberty and order. Instead, government has simply become the vehicle whereby the Hobbesian war of all against all is carried out. Regulation becomes its means as regulators become agents of predatory interests.
Is there evidence that more competent government leads to less burdensome regulation?
Let's take track and trace. Those governments that have been best at track and trace have had the most burdensome COVID interventions. The most lockdowns. The harshest impositions.
I'm not just referring to China here. Would any of us want the situation they have in Australia?
Even if we could have a government that was good at what it did, that wouldn't mean that "what they did" would be good. We might be better off with a government so incompetent that nobody wanted it to do anything. That seems to have been the case with COVID. Let it rip was by far the optimal solution with no close second.
I don't think competence is the major problem with government. It's that governments don't have credible signals of just what they should be doing to make peoples lives better. Capitalists have price signals. Governments have elections. One is a lot better than the other. It's not clear that taking away elections would improve the governments ability to figure out what it should be doing. Unelected Public Health Officials for instance have had way worse instincts than politicians.
Yes, Arnold, you can bet that all framers of past liberal constitutions never anticipated the world we live today. If you were familiar with the history of liberal constitutions in Spain, Argentina, and Chile, you would fear greatly how political competition for grabbing and holding power is being played today. Anyway, I think today's U.S. political competition is much worst than in each of those three countries for the simple reason that you have never faced the high degree of corruption of all liberal values and institutions I see in your country today. Too many have remained silent for too long in your country -- the only one with a good, long record of liberal values and institutions (yes, the poor record of those values and institutions in the last 200 years is a common excuse some intellectuals use to explain what is going on in Spain, Argentina, and Chile).
Regarding your proposal to reform just the bureaucracy, let me remember you that to reform radically the bureaucracy requires first a substantial reform of the Constitution -- but, as I used to discuss with Gordon Tullock in early 1980, reforming the Constitution at best is a necessary condition never a sufficient one.
Name a single governmental organization that was good, declined in quality, was reformed, and then performed well again.
https://www.washingtonexaminer.com/restoring-america/fairness-justice/the-death-of-democracy
This essay from November notes real problems; your ideas would likely reduce such problems. So I like them. I was really glad to see you note a couple of places the need for some term limitation, like 6 years. But don't believe they're realistic in the near term.
The Deep State "Fourth" branch needs term limits throughout. "Public Service" should be very limited for all. I'd propose that after 10 years, no more raises nor promotions (tho same-pay relocations OK), and conversion to a 1 year contract. Which can be renewed at most 1 time.
The "Civil Service" needs big reform - it ain't gonna happen.
Small, but very significant reform seems possible - like bureaucrat term limits. The single key metric for gov't agencies should be turnover - it should be 1% MORE than non-gov't wage earners (not small business owners). General IRS based turnover, reporting primary income from a different source this year than from the year before, should be a key number. Like 12% (or 10% or 15% or whatever it is).
Moving most gov't employees out of DC would also be generally good, and seems more politically possible.
I strongly fear that your ideas will be a case of "The Perfect being an Enemy of the Good".
I would support politicians supporting your reforms - would you support other politicians supporting other reforms that might make the situation less bad (/better ?) tho not as much as you propose here?
Alright, I wrote up some problems with your proposal, and added some suggestions for improvement. (https://dochammer.substack.com/p/state-capacity-libertarianism-again)
"Pick any problem the government is supposed to solve, and you will find the same problem: how do you measure success, and how do you identify how well the agencies should be doing?" is I think the key question. Governments lack the competition that markets provide, and so lack the mechanism for generating knowledge of how well the organization could be doing. Without being able to point at another org and say "Look, they manage to do X. Why can't you?" there is nothing to drive estimates of whether your division is doing well or not, other than just the vague feelings that drive our politics now. The COO and CA proposal does not contain a way to improve the status quo.
You say the administrative state is necessary to provide order; I say it hasn't been shown that the administrative state is even capable of providing order, much less necessary.
I also think that your proposal might fly if we were move the federal regulatory systems down to the state level, as at least there is a little more competition and comparison there. I don't think that solves the fundamental problem, however, and certainly it isn't the way current political currents flow.
Arnold;
I would like to point out something that is missing - you start with twin purposes for government in your first sentences - order and the promotion of virtue - and then proceed to only discuss the requirements of order and say nothing of virtue. Could you consider addressing it?
In order to create real accountability, your CA would likely need the power to fire individuals and close entire agencies. In fact, it might be best if the CA organization was required to terminate whoever it thought were the worst performing 5% of staff each year so that they can't dodge their responsibility for identifying them.
Idea in progress? Federal gov does have a GAO and aren't the Cabinet Secretaries supposed to be COOs? The questions seems to be not about creating new positions, but...(1) increasing the independence and effectiveness of the GAO and (2) making the huge Federal bureaucracies more transparent to Cabinet Secretary management. Problem: no one in the bureaucracies, the entities subject to regulation, nor Congress want either (1) or (2).
Arnold, the centerpiece of your proposal is purposefully putting successful businesspeople into high positions of government. In the recent past, what business figures who assumed roles in the executive branch have you most admired?