Featured Post

The Dangers of Complexity in Policy Formation

On Zic’s piece on bodily integrity, Michael Drew and I had a disagreement about one of the things that really annoys me about the US healthcare model – the way it treats private companies as if they were public utilities – simply directing 3rd parties to enact government policy, rather than doing it themselves. This is what I had to say in regard to the Hobby Lobby case that was the focus of Zic’s piece:

This is why the whole premise of the US health care system (a premise that was only further entrenched by the ACA) is flawed – that employers have any business acting as healthcare providers. It is only because of this policy-induced intermediation of health care services that the Greens are even a party to this transaction. A single-payer system wouldn’t have this problem, nor would the individual-insurance-plus-welfare model I prefer.

Michael wanted me to justify my position:

Aren’t some very successful health care models based exactly on the utilization of private utility-like organizations that are regulated to within an inch of their lives? (Germany, if memory serves?) I personally want that kind of thing on the table, which is not of course an argument that I want to adopt it in any given case. As a general matter, I hardly see where I should reject that out of hand. If you want me to think I should, either as general principle or a specific case, you’d have to make the argument. (That’s not an invitation to do so: if you want to, go for it. But my point is just that the truth of either the specific admonition you made nor of its general form is self-evident. As it stands, it’s just simple, unsupported normative assertion. “Things should be this way! Because!”)

This is a fair point, I didn’t state my case. So, here is a quick run-down of the reasons why I strongly prefer more direct government intervention of policy by proxy:

1) Simplicity: Simplicity counts for a lot more than people give it credit for. People have limited cognitive resources, and making a policy too complicated can make it too hard to deal with. This is not to say people are stupid, it’s just that dealing with something you don’t do often can require a lot of attention and effort, and every little thing you make people keep track of makes their life that much harder. The same applies at the government end, the more complex a policy is the more likely there will be a failure – either a mistake at the policy level, or some kind of flaw in the operational policy that makes the policy fail to work as desired. The more thing your plan relies on going right, the more likely you are to have something go wrong, or in the immortal words of Montgomery “Scotty” Scott: “The more they overthink the plumbing, the easier it is to stop up the drain”.

2) Distortion: Distortion is the economists’ term for a market going wrong because of bad incentives. In practice what it means is that people end up doing things that are destructive, or at least unhelpful, because there are incentives that make it worth doing for them. The more you concentrate the costs of a policy onto a small group, the more perversely they are likely to behave. I strongly suspect a lot of what is going wrong with U.S. healthcare is that the risk-pooling requirements and coverage mandates imposed on heath insurers lead the insures to want to drive away their sicker clients. Distortion is something to be avoided wherever possible in policy making, and where avoiding it is impossible it would be minimised.

3) Disintermediation: Consumer choice works best when there are a few steps as possible between consumer and the producer of a product or service. Health insurance is already one step removed because you have insurers making decisions on behalf of customers, but that’s unavoidable because of the risks involved in health care costs. But having employers decide what insurers will make decisions for consumers is an extra, and unnecessary, step. At least if insurers are being paid directly by consumers, they have to worry about consumers shopping around for another insurer. But how many people would quit their jobs because the health insurer is irritating to deal with? Every extra step in the decision-making process is a little more grit in the gears of the market.

4) Pluralism: The Hobby Lobby case happened because some people feel the requirements of the ACA go against their moral principles. Depending on how the court rules, one of two things will happen: Either a hole will open up in the protections the ACA is trying to afford to people, or the opposition to the ACA will grow more heated. Neither of these is good for the ACA in the long run. Now you may ask, why should I care about the Greens and their values? First off, because so long as they have some ability to affect policy, their opinion matters in a Realpolitik sense. And second, because policy pluralism is designed to stop moral conflicts escalating into political, or worse military, conflicts. Pluralism means making some concessions to people whose values you despise, and making people pay taxes of things they object to is less of a burden on their conscience than actually making them do something they object to – it’s like the difference between paying for an army whose actions you oppose, and actually being drafted into that army.

5) Obligation: Typically, the argument in favour of providing social support is that society has an obligation to ensure the most vulnerable do not suffer (at least up to point). If that is the rationale for social assistance, then why push the burden of this obligation onto a subset of the population? To use conscription again as an example, conscripting soldiers places a disproportionate burden on a subset of the population (usually young men), while a taxation-funded professional army spreads the burden far more evenly. Admittedly, this point applies less to the issue of health insurance (since workers most likely bear the majority of insurance costs), but this point applies in other cases.

6) Accountability: This is probably the biggest issue for me. The difference between a modern democratic government and a tyranny, is that modern governments have rules. There are rules about government actions being disclosable, rules about government spending being accounted for and rules about how governments may act in regards to the public. The specifics of these rules, and how well they are enforced in practice very from country to country, but none of them apply to the private sector. If a policy is funded from taxation, then that costs is clear and known, but who knows how much the ACA will cost? Only the government’s part gets counted, and much of the action as been foisted onto the private sector. Without Transparency the government cannot be held to account by the public. How can they possibly decide if a policy was a good idea, if its costs are largely unknown?

Ultimately, these forms of pseudo-private provision often result in the worst of both worlds. They lack the price-responsiveness and innovativeness of markets, but often lack the comprehensiveness of public provision.

So what should governments do instead? My suggestion is to either go more public and/or more private – I say and because sometimes its about refocusing the government on the aspects of the problems that markets address poorly, and in other respects letting the market work. This is why I prefer a healthcare approach consisting of private, individual insurance backed by welfare because it lets the market do its thing, while focusing the government on the problem the market can’t solve – that some people can’t afford their expected lifetime healthcare costs.

Please do be so kind as to share this post.

94 thoughts on “The Dangers of Complexity in Policy Formation

  1. I don’t think there is any liberal who would disagree with you and possibly not any conservative. I think there is near universal consensus that it is extremely problematic that many to most Americans get their insurance through their employer*. France seems to have some kind of hybrid market where there is a base national insurance and then supplemental insurance through your employer or purchased privately. I don’t know too much about the system.

    The problem is that American liberals would most likely want a system that is like NHS or Medicare for All and American conservatives would want something extremely more private. No one will touch Medicare for the elderly because that is a sacred cow in American politics and no one wants to piss off seniors.

    I think ACA tries to inject some consumer choice into the market but I did not personally see much competition on the healthcare exchanges. The plans are labeled gold, silver, and bronze. All the bronze plans were very similar, all the silver plans were very similar, all the gold plans were very similar, etc. The difference in pricing was insubstantial. There wasn’t anyone trying to sell me a silver plan at 200 dollars or a gold plan at 300 dollars. Note: I currently pay about 353 dollars a month for a silver plan for one person.

    *That being said as someone who has been freelancing for the last two years because of the sheer luck of graduating into a collapsed market/recession, I am yearning for a full-time long term jobs with benefits including health insurance but mainly paid vacations. I do not buy the libertarian argument that these are things I can bargain for as a freelancer (see: asymetric power relationships). Many of my friends refer to jobs with regular salary and benefits as “real jobs” because they are the jobs our parents had when we were kids and were part of the social contract of our youth. Even if there was some kind of national vacation policy, I am sure freelancers would be excluded. Paid vacation/PTO makes travel easier because you know you have to save for the vacation only. You don’t need to save for the vacation and rent and other expenses while not working, etc.


  2. , this sounds like an argument for a single-payer system as opposed to a nearly completely-privatized one.

    Simplicity: Single payer is a whole lot more simple because there is only one actor involved in deciding what health care to pay for. The decision about what is necessary is made between doctor and patient, and then the government says that yes, it’ll pay for it, or no, it won’t. A private insurance company acting in concert with a government actor like Medicaid adds a second decision-maker, and in a system where the patient is also expected to make a contribution, that adds a third actor who might throw money in the pot.

    Distortion: This would be minimized because the costs of payment would be rendered as diffuse as the tax base. Sarah Palin’s feared “death panel” would come in to play as a decision choke-point, as would the formulation of payment-per-procedure tables. But these seem like they would be the only places where distortion could come in to play, and both would be subject to democratic checks.

    Disintermediation: As above, the decision about what care was needed would be made between doctor and patient, with no involvement from any insurer. (Query the extent to which insurers actually inject themselves into these decisions today, in non-extreme cases.) If the government won’t pay, then doctor and patient have to come up with a plan B.

    Pluralism: Here, that’s the Hobby Lobby problem — and the Greens have no standing to object to the government spending tax money on health care needs in a single-payer system any more than they have standing now to it spending tax money on other things they might not like such as war or foreign aid. Nor would the Greens be directing their company to spend money that would otherwise go to them on products to which they have a moral objection.

    Obligation: Again, the obligation to provide for medical care becomes as diffuse as the tax base.

    Accountability: This is the area where there is the most fear, but there need not be in a democracy, as eventually everything falls under democratic control, albeit some of it in an indirect fashion. Almost certainly in the U.S. a single payer system would be overseen by Presidential appointees to an executive agency, possibly with the advice and consent of the Senate. It would look something like the Federal Trade Commission or the Federal Communications Commission, with a group of commissioners serving rotating, staggered terms, presiding over regulatory decisions made by subject matter experts (aka bureaucrats) about what money should be paid for what service in what market.

    Seems to fit a whole lot better than more aggressive privatization, particularly given that as comparators go, the U.S. is already one of the most privatized health care delivery systems in the industrialized world.


    • Single payer is still inferior to a simple cash transfer, in terms of efficiency. Also, reforming a single-payer system is likely to be harder than consumers (some of them subsidized by cash payments) voting with their feet in a market-based system, where innovations can happen whenever someone has a better idea.


      • I’m certain there’s a bunch of studies I can cite about how humans aren’t terribly good at understanding small risks, and understanding why insurance is necessary.

        Somehow (through laws, I believe) we’ve managed to inculcate the idea of “car insurance”. Through mortgages, on the main, our beloved free market has managed to inculcate housing insurance.

        Would we be giving a cash transfer for insurance, or for health care?


      • I think you are assuming that individual consumers are sophisticated enough and armed with sufficient information to move their dollars to good innovations. IMO, healthcare is significantly more complicated than regular-versus-organic lettuce.

        That doesn’t mean your theory is wrong, necessarily, but the market responsiveness is going to be dampened by brand loyalty or inertia, feel-good advertisements, and competitive pricing schemes that sacrifice quality innovations. We see this now with choice of banking institutions (inertia), insurance and financial services (influence of effectively content-free mass media), and air travel (most consumers value price above quality, obscuring quality differences between competitors).

        describes a ranking of the quality of various products as “Gold,” “Silver,” or “Bronze.” That’s probably simple enough that a consumer can understand where the extra money is going and make a rational choice. But digging down further gets real complicated, real quickly. So I’m not confident that consumers are equipped to make choices with the level of information necessary to achieve the sort of market-sorting that aims at.


      • I’m not sure how complicated it is.

        Do you want Dental? For how many? The price to you is x. These are your benefits/caps/deductibles
        Do you want Eyecare? For how many? The price is y. These are your benefits/caps/deductibles
        Do you want Medical care? For how many? These are the plans to choose from. Here is a table showing your costs, what the plan covers, the prescription plan, your caps/deductibles, etc.

        It takes more time than doing your taxes if you really look at what you’re getting and optimizing your policy, but it’s not THAT hard. I reviewed mine and actually made changes in my plans because of some changes (thanks obama) and it took me, maybe an hour, for the medical. Everything else was simpler.


      • Damon,
        Did you look up statistics about denying care?
        Before this regulation, there were a lot of dirty insurance companies that really would take your money and refuse care (“preexisting condition!”) on any pretext they could find.

        If everyone was being honest, it would be mostly easy — but you’d still want to look up statistics on “how good is the care” within individual networks…


      • Dental is easy, . It’s one of the models that actuaries train with. There’s only so many things that can go wrong with the mouth and a relatively small number of risk factors predict them with high accuracy.

        Medical is just plain more complicated from an actuarial perspective. Too many things you can’t predict, too many variables.


      • Burt,
        That’s actually cost shifting, I’m pretty sure.Having something go wrong with your mouth can lead to massive infection (entire face swelling up), or septicemia, or a lot of things that the Health Care Docs get to deal with (including heart attacks).


      • Burt,
        Maybe we’re talking past each other. I’m talking about your comments to Saul. I’m not talking about actuarial stuff, but the difficulty of consumers to select a plan. As an example, I had a grid listing each plan available and the relevant costs/benefits, etc. available. Links were provded to each Heath plan’s site so I could look for doctors, etc. Maybe if you’re offered a dozen or so plans it could get tedious, but I have 4 plans to choose from.


      • I think that a lot of people who thought they had “great” insurance only thought so because it reliably paid for whatever piddling costs they ran up during day-to-day medical procedures. Most of us are lucky enough not to have to really test our health insurance and see where the coverage gaps, loopholes, and limitations really are.

        I can rattle off my deductible and all of the other basic parameters of my health plan and I understand the basics of insurance better than the average Joe, but if you ask me whether a particular procedure is really “covered” with a particular provider the best answer I can give you is, “I hope so.” I don’t think I’d reliably be able to distinguish between plans at that level of granularity, and I honestly have no idea whether I’m getting a “good deal” when I need anything more than a physical.


      • Merely defining what counts as medical care and what doesn’t is extremely complicated. Does Viagra count? Birth control? Arthritis medication? A cardiac stent? Different people will have different ideas about what should be covered, so this will have to be spelled out in painstaking detail. At which point, it becomes so complex that it’s hard for consumers to make an informed choice. I have no idea if I’ll end up with rheumatism, or what treatment options will work for me if I do get it. To expect me to choose an insurance plan on the basis of what it covers, today and with minimal medical knowledge, is an enormous ask, well above the reach of many consumers.


      • Go pick me out 10,000 Americans.

        Then survey them all with these questions:

        * What ten medical conditions are you most likely to suffer from, that will cause an economic crisis for you if they occurred in this calendar year, given your current income, expenses, and savings?

        * How likely are you to contract any of those conditions, and how much of the cost of treating those conditions does your current insurance provider cover?

        * Given all of that, what is the expected value of your insurance to you, this year?

        Now, if you had to hazard a guess, how many of those 10,000 Americans would be able to answer those questions, and of those, how many of them would get their answers anywhere near the actual real answers?

        I have a guess as to how many, myself. Maybe you have a different answer than I do. That guess leads me to believe that the information asymmetry problems here are kind of immense.

        That sorta makes the market solution for insurance a problem, as far as efficiency goes.

        Simple cash transfer might be more efficient than having insurance in the first place, but given the savings rate in America, and the inability to amortize health care costs when you incur them, that would be… problematic.


      • Given your questions, I’d agree they’d get the answers mostly wrong. I’d then point out that the largest source of such misinformation is the popular media (COUGH! HuffingtonPost COUGH!), and that people aren’t dumb, just far too trusting of such outlets.

        I’m with James, a government information campaign would be more useful. And more in line with functions government should be trying first.


      • Why do we assume the solution to an information problem is a substantive government program, rather than a government information program?

        Why would you assume that I’m assuming that? I was just pointing out that there was something of an information problem there. Certainly there are ways to solve that problem.


        I’m with James, a government information campaign would be more useful.

        Our existing government information campaign system is underfunded, rather terribly complex, and a political football, and that’s just for the kiddos. Remember the brouhaha over “death panels”?

        Anywho, I think the average math skills of the average Americano are pretty much crap. And while it’s certainly the case that a large number of them could get better math skills – or we could automate the presentation of those skills, so that they could just get the info without the k-nowledge, there’s a mountain of psych lit that says that they won’t listen to the info if it doesn’t conform to their expectations, anyway.

        I dunno. Forty years of antismoking campaigns, regressive taxes… and 18.1% of the adult population still smokes.


      • Funds can be re-allocated easily enough.

        The population of smokers has been dropping for years.

        there’s a mountain of psych lit that says that they won’t listen to the info if it doesn’t conform to their expectations, anyway.

        Yeah, I got no easy fix for that.


  3. I completely agree about greater simplicity being better. Regarding HC it would have been great if the German/Swiss model of highly regualted insurance companies providing coverage to all, without an employer being involved, and optional add coverages being available. Great ideas to think about. Unfortunately looking at the experience of other countries was pretty much shot down by the conservative side here.

    Inertia or fear of change or status quo bias, whatever we want to call it, is a powerful force here. Most people wanted some change to HC but a lot of those people only wanted it as long as it didn’t affect them. This was of course exacerbated by the scare tactics used to fight against any HC reform. Many liberals fell into the trap of thinking we would get a massive rework of the system when it was far more likly we would get half a loaf at best. In fact many of the critics of the ACA sort of do that to. They criticize the ACA for not being a massive change yet that massive change could never happen given all the veto points in the system and the complexity of the system being remodeled. I agree with many of the realistic criticisms of the ACA but am more hopeful because it has started the ball moving on change. Not all of the changes are good, but without some movement, nothing was ever going to happen. Given that the best reform we could get was tinkering and nudging what we have in a different direction and to solve some big problems we were never going to get something simpler.


  4. When you impose wage and price control in an inflationary market (WWII), you get all kinds of suboptimal results. Linking healthcare with employement as a way to give employees add’l compensation was one result. The tax breaks companies got further entrenched this.

    The Nazis are still effecting our internal polices 50 years later.


  5. One can often cast the same policy as ‘government mandate’, ‘public-private partnership’, or ‘corporate giveaway’ and go around the political horn in the support/opposition split.


  6. For payers, the central question is “Who is going to pay and how much?”

    In the light of that question, complexity is a feature, not a bug.


  7. Simplicity may be the wrong word. May I propose elegance as a substitute. Take for instance the human eye, one of the perpetual props in the back and forth between natural selection and creationism/intelligent design. The eye is extremely complicated, but it works, because it evolved over lots and lots of iterations. It’s the result of a process that had lots and lots of dead ends.

    The problem with government solutions is that they are almost never elegant, because elegance almost never comes straight from the drawing board. Government solutions are mostly the result of a combination political compromise and technocratic planning. And the process does not really allow for multiple iterations of trial and error. We basically get one shot to pass a new entitlement program and then we are stuck with a system that is legally mandated to remain the same. Reforms happen, but not at the rate that they ought to happen.


    • I rather think ACA has some very elegant components.

      Take the contraceptive mandate. The law has a simple passage — provide basic preventative care without a co-pay, and direction to specifically study basic-preventive care for women. The studies that show contraception is critical basic health care for women are legion. That’s a very simple, elegant, and direct way to address a preventible and life-changing health-care concern that will impact one out of every five women.

      I am also impressed with how the subsidies function; subsidizing income through the IRS (the agency that handles income, btw) so that the cost of their health care will not force them below the poverty line. It may force you to it, but not below it.

      I am very anxious to learn more about how cost-efficiency studies change how we access care. Already, we’ve seen public warnings on breast and prostate cancer.

      So I don’t think you can discuss the elegance or lack of elegance in ACA, certainly the most complex domestic government policy we’ve had since Homeland Security, on the healthcare.gov rollout or early enrollment. Those are new, the changes to the medical system were put in place first, and we’re just beginning to see how that will look.

      There are flaws. Huge flaws. But elegance, too.


      • Thousands upon thousands of pages of elegance creating myriads of administrative and regulatory agencies.

        I suspect you two are not are not using the term in the same way.


      • Yah, can’t call it elegant. The IRS has no enforcement mechanism regarding that tax. Any mechanism likely to be put in place will likely be another massive invasion of our personal information. The subsidies and industry payments are set up to hide true costs, there’s no history for the rates and they will likely rise in the future, leading to top down cost controls being imposed.

        I don’t see that as elegant. I see it as a tangle of briars that will be difficult to untangle, will be painful, and will take a lot of time.


      • I specifically referred to the subsidies and you responded with enforcement, which would be for people who opt not to purchase insurance (or cannot receive Medicaid because the governor of their state opted out). So you never actually bothered to address what I considered an elegant handling.

        Kaiser has a subsidy calculator with which you can explore it.


      • Actually my comments applied to both the enforcment mechanism and the subsidies. I said “The subsidies and industry payments are set up to hide true costs, there’s no history for the rates and they will likely rise in the future, leading to top down cost controls being imposed.

        I don’t see that as elegant. I see it as a tangle of briars that will be difficult to untangle, will be painful, and will take a lot of time.”

        How does that not address your point about the elegance of the subsidies?


    • Plus one for elegance.

      Mark Pennington suggests another alternative… “Robust.” He is alluding to the same features of iterative learning, constructive competition and success in real world rather than ideal world conditions.

      And though I haven’t read the book, from the various articles I have seen, I am pretty sure Taleb hits the same idea but refers to it as Antifragile.


    • Or take the human throat, which has evolved over billions of years to elegantly combine the functions of breathing and eating, allowing us to choke to death on small bits of food.


    • :

      Simplicity may be the wrong word. May I propose elegance as a substitute. Take for instance the human eye, one of the perpetual props in the back and forth between natural selection and creationism/intelligent design. The eye is extremely complicated, but it works, because it evolved over lots and lots of iterations. It’s the result of a process that had lots and lots of dead ends.
      The problem with government solutions is that they are almost never elegant, because elegance almost never comes straight from the drawing board. Government solutions are mostly the result of a combination political compromise and technocratic planning. And the process does not really allow for multiple iterations of trial and error. We basically get one shot to pass a new entitlement program and then we are stuck with a system that is legally mandated to remain the same. Reforms happen, but not at the rate that they ought to happen.

      That’s an interesting metaphor, and perhaps on more levels than you intend.

      The human eye is always held up by various people (usually social conservatives in my experience, but certainly others as well) as being the perfect, elegant solution of evolution or the creator. But it isn’t. It’s surprisingly susceptible to trickery, it can’t really see in the dark, it’s both delicate and exposed, it can’t “stand guard” while you’re sleeping the way your ears can, it’s both overly complicated and inside your head which means you need very advanced technology to make the most basic of repairs to it, and for the majority of people it doesn’t work properly when looking at thing up close, at a distance, or both.

      Today’s conservatives and libertarians look at the human eye and ask themselves, “what about this works the way I want it to,” and declare it an elegant success. Similarly, they look at actual and potential government solutions and ask themselves, “what about this doesn’t work perfectly,” and declare them failures.


      • Good points, Tod. Additionally, do you know why we have a blind spot? It’s because our retina is constructed backwards. It has layers of light sensitive cells lying atop a layer of supporting cells. Good enough. Except the support cells are in front of the light sensing cells. Light has to travel through the support structure to reach the rods and cones. This also puts the blood vessels, capillaries, and optic nerve fibers on top.

        Other animals, whose eyes independently arose through different evolutionary pathways have this all flipped around the way any half-way competent engineer would design it.

        Don’t even get me started on the vagus nerve…


      • You guys are making great points on eyes (and even elegance) and still missing what I see as the key insight from jr. Indeed, as I suggested elsewhere, I don’t think elegance (or efficiency) is a good word for the concept either and so I pointed out two other terms.

        When you read how jr describes it he is defending an evolutionary adaptive system which gains from trial an error and iterative learning. He is describing a complex adaptive learning system.

        Your final comment on how some look at “what about this doesn’t work perfectly” and declare government “solutions” failures, misses this dynamic. Much of the differences in frameworks or paradigms between the left and right is not just over the what, but the how.

        In my words, choosing to try to solve problems via government may be the efficient way to solve the problem for some classes of problems. Indeed I think some Public Choice theorists would justify government on grounds of efficiency. Other classes of problems are best solved via experimentation, iterative trial and error, learning from mistakes and so on. An example of this type of complex adaptive learning system is evolution. Another example is the scientific method. Another is markets.

        All four systems have their place. Often the essence of a disagreement from the right is that they believe we are using the wrong problem solving system. I

        n addition, there are proven ways to make government problem solving more adaptive where appropriate. Subsidiarity, choice, opt outs, constructive competition between ideas and organizations, sunset provisions, trial programs, and so on can all be appropriate for some circumstances.


      • Roger,
        there are half a dozen ways you can construct decent adaptive systems within government. If the republicans were honestly suggesting decent alternative solutions to Obamacare, I’d love to see them in action too.


      • Roger,
        see, this is what I hate about binary perspectives. It says that liberals can’t like a decent solution that is based on the free market, even if they want a better solution based in the government. (speaking specifically of global warming solutions).


      • As Roger said, I think that you are missing the point of the example. No, the eye is not perfect, but we are not looking for perfect policy solutions. We are looking for solutions that do what they are intended to do without expending an unreasonably amount of resources and without causing all manner of negative unintended consequences. And those are two areas where government planning tends to be really terrible.

        And government failures pretty much speak for themselves. I don’t need to get terribly creative to find fault with the War on Drugs or the invasion of Iraq.


  8. A great post, James. I think each of these issues tends to get overlooked in the desire to just finally get something done. Maybe if it was easier in our system to get something done, it would be easier to focus on getting something done right.


    • Maybe if it was easier in our system to get something done, it would be easier to focus on getting something done right.

      Very much this. It is one thing in a parliamentary democracy for the government to present a coherent program that ticks the various boxes outlined in the original post – it also helps that a program would be the platform that the government was elected upon. The whole ship of state is steered by a captain that can give directions. Watching a Queen’s speech in the UK is a universe apart from watching POTUS deliver the state of the union. “We should do this, that and the other” in the US. “My government will bring forward a bill that will do XYZ” in the UK.

      Unfortunately for the US, the sainted founders were all about competition between the branches. I mean, we have a SCOTUS case about when is the Senate in recess or not, whether or not the POTUS can issue passports with Jerusalem on them in defiance of a congressional claim to command that passports say Israel… and on and on, interbranch conflict and disagreement about how far precisely their powers go. We don’t even have clarity on how warmaking happens – can the president legitimately initiate? when? does passing a spending bill indicate congressional acquiescence? – another pretty significant thing that other systems can pretty clearly comprehend is a contest of competing claims to power in the US.

      The US system isn’t prone to efficiency, or elegance, or delivering the coherent program. Would that it were capable of such efficiency.


    • James — this is a fantastic post.

      You have echoed–but in more concise and elegant terms–a thought that I have been formulating for a couple years now. Simple policies work better than wonky ones.

      During the year-long debates around creating Obamacare, in 2009, I had the opportunity to talk with many people about it. Since I live in Orange County, a good portion of those many people were conservative. Interestingly (to me), the biggest source of objection to Obamacare was its complexity (followed closely by the purchase mandate). Among those that I spoke with, there would have been less objection to a single-payer system than what ended up emerging from Congress–even if it would have resulted in higher taxes.

      I think that it’s one of the core characteristics of Americans to distrust government (and concentrated power generally). When policies are complex, people tend to believe that they’re getting screwed. So you can have an army of policy analysts try to carefully align interests and incentives so that the “public good” triumphs: but complicated policies always end up creating dark corners and twisted alleys, where the clever and well-lawyered can subvert their intent, and perverse incentives arise from unanticipated behaviors.

      In fact, Roosevelt’s real genius was creating and sustaining competing interests that would advocate for themselves, over static and rigid constructions of overlapping regulation that never quite map human behaviors adequately.

      This is, I think, the key insight that FDR had that all the wonky neo-liberal Democrats from Jimmy Carter on have really failed to understand. People supported Social Security because it was easy to understand, and seemed fair to them. And his administration’s straightforward support of unionization helped produce the economic compression of the 40s through the 60s that made the US the first truly middle-class society on earth.


      • That’s a selective and survivor bias reading of FDR’s administration, methinks. His presidency loved some wonky complicated solutions (I mean, they were called the Brain trust, first somewhat derisively, but then they made the term their own).

        Many, many aspects of the New Deal were abandoned, well before the Carter/Kennedy deregulation Mothra and the Reganomic Godzilla started smashing things – some due to Court rulings, some because they got rolled into wartime measures (which then were phased out under Truman), and some because even their advocates at the time saw they didn’t work.

        Social security, has, of course, lasted to this day, and a good part of the reason is indeed, that it is simple. (and now has lots of stakeholders). But it took a course correction in the 80s that brought in a significant sector of the economy previously outside it (most public sector workers), plus, since then, the fading into history of the Railroad Retirement system. (as a competing claim)


      • Many, many aspects of the New Deal were abandoned, well before the Carter/Kennedy deregulation

        This is an under-appreciated point about the New Deal, it was not a cohesive whole, but rather a group of policies, each of which should be evaluated independently.


    • I think I have said this before, but I conflated difficult to get things done with gridlock. I wonder how US history would have turned out if its founding fathers had been a bit less radical and instituted a parliamentary democracy closer to the British model (i.e. if they had been more like Canada).


      • I’m not sure radical is quite the right word there. Madison actually proposed a parliamentary system, with the executive being selected by the legislature (although he didn’t specify that it would be from within the legislature; that was left open). His plan can be read here.

        That was almost accepted, until it was noted that that the plan didn’t have separation of powers, which as every good English subject–which they had been until recently–who’d read Montesquieu knew was crucial to good government. Gouverneur Morris objected that the selection of the executive would “be the work of intrigue, of cabal, and of faction…”

        His proposal for direct election of the executive was rejected as too radical–George Mason argued that it “would be as unnatural to permit the people to elect a President as it would be to refer a trial of colors to a blind man”–and they compromised on the electoral college.

        Both sides were bound by their vision of the British system as the best in the world. They had never objected to it, just to the refusal to let them participate in it. And they were deeply influenced by Montesquieu’s misinterpretation of that system. Their only real radicalism was the rejection of a monarch, which left them puzzling over how to properly create the executive power.


      • The parliamentary system was still in embryonic form at the time the Constitution was being written. As James Hanley pointed out, the Founders were operating on their own interpretation of the British system, which was kind of out of date at that point. Besides Madison, they wanted something like how they imagined England operated after the Glorious Revolution with a President instead of King and a Senate rather than a House of Lords.


      • is right, in the 1780s executive authority was still in the hands of the King and the Reform Acts hadn’t been passed yet.

        Honestly, I think you can attribute almost all the mistakes the Founding Fathers made to the fact that they were doing something that had never been done before – it’s almost impossible to get it right in those conditions.


      • It puzzles me how people can be so systematically mis-informed about the structure of their government and still run a successful revolution. Surely some among them must have been to England, come back and said: “This Montesquieu guy is full of shit”. Back in the day when not every constituency was contested (because there weren’t enough opposition candidates) the average Singaporean still knew that every member of the cabinet was the MP for some constituency. I’m not saying that such knowledge should have been common in the days before television, but surely those who belonged to the educated classes (as the American founding fathers did) should have known better.


      • , the Founders weren’t mistaken. The parliamentary system as we know it today did not exist back than. King George III still had some very real powers and the link between the Prime Minister, Cabinet, and the House of Commons was tenuous at best. The title Prime Minister was used but looked down upon during the 18th century. It wasn’t mentioned officially in legislation until 1917. William Pitt, Prime Minister at the time, was appointed despite Parliament not because of it.


      • Didn’t Madison’s plan also have a problem for the small states? “Quota of their contribution” and all that? Of course, bicameralism (ie a senate-like second chamber) would have taken care of that, which is how things were likely to turn out in the end regardless of whether it was a parliamentary or presidential system.


    • A great post, James. I think each of these issues tends to get overlooked in the desire to just finally get something done. Maybe if it was easier in our system to get something done, it would be easier to focus on getting something done right.

      I admit that when the ACA was just a twinkle in Baucus’s eye, I was of the mindset that this was the one best chance to get widespread coverage of some sort, and so I was prepared to support almost anything. Now I’ll have to eat crow if in 5 or 10 years’ time, premiums skyrocket or other perverse things happen and if instead of somehow improving the system along some lines, we’ll return to the status quo ante ACA.

      Having said that and with due acknowledgment that none of us knows the future or the consequences of counterfactuals, I don’t think I was far wrong to say that it was unlikely we’d get another chance to do something. So one question is whether the current system is worse or better than it was before? It’s probably better in some ways and worse in others.

      Another question is, should we tinker so radically with the way things are or should we try to change things more incrementally? One answer is that the ACA is in some ways more incremental than, say, a single payer or even a “robust public option” because it worked on, and in so doing further ensconced, features of the existing system.

      Another answer is that any proposed change could go wrong. If I’m not mistaken, you have in other places expressed support for something like a freer market for most health services, but with government support for those catastrophic or chronic cases or cases of indigent persons. (I forget where I read that, and if I’m making this up, please forgive me.) If such a plan had been proposed by someone in Congress, what would have been the resulting legislation? What exceptions would have been worked in to get past the various veto points? I don’t know, but I do think there would be enormous pressure from some employer groups to retain the tax incentive for insurance (others might go the other way, of course), insurance companies would want to keep their antitrust exemption and some of them would want to retain the ban on interstate insurance sales, and enough fiscal conservatives would likely work to make the catastrophic/chronic/indigent government plan as lean as possible. Whether and to what degree these interests would succeed, I don’t know. But that is something to consider.

      I know I’m saying that the perfect is the enemy of the good. And maybe there is less comprehensive change that could plausibly be pursued and stand a chance to be adopted. And even if there isn’t, that oughtn’t silence discussion as to what the best policy should be. But in the meantime, there are problems that need to be addressed, and we have to sometimes work with the government and veto points and vested interests that we have and not the ones we want.


  9. Sometimes you need a complex solution when simple doesn’t work or isn’t feasible.

    The key to complex solutions is openness to experiment, empirical review by experts of experiments, the acceptance that solutions will need revision as problems change, and copying what works elsewhere.

    Public programs can do that. The reason we don’t see experiment, review, and revise this in our healthcare system is because when many Republicans get the experiments or revisions they initially suggested, or anything short of an extremely right-wing proposal that has never been tried, they scream death panels and use the choke points of the legislative system (that even a minority can control) to kill any small bore reform.

    This is obvious to all.


    • Sometimes you need a complex solution when simple doesn’t work or isn’t feasible.

      Indeed, sometime the best solution is complex – complexity is an evil, albeit sometimes a necessary one.

      The key to complex solutions is openness to experiment, empirical review by experts of experiments, the acceptance that solutions will need revision as problems change, and copying what works elsewhere.

      That is best practice, yes.

      Public programs can do that.

      Theoretically, yes.

      The reason we don’t see experiment, review, and revise this in our healthcare system is because … Republicans

      Even in parts of the world that don’t have Republicans, governments still have trouble learning from mistakes. Tim Harford in his book Adapt, makes what I feel is the best case for why that is so – voters hate leaders who admit they made a mistake. That creates pressure for politicians to avoid adding falsifiability to government programmes.


      • I think it’s much simpler: looking for mistakes and evaluating programs requires funding; since it’s not direct, it’s easy to cut or never budget in the first place. We lack a habit of evaluating.

        This is a problem akin to another that results in bad policy, average citizens focus on the legislative process, but that’s only the first step in an ongoing process. Rule making, where legislation is turned into rules for bureaucrats and citizens to use, is where the bulk of regulatory capture happens, and where most people stop paying attention.


      • I disagree with this. Rule-making bodies implement public process; it has the exact same requirements for notification, public hearing, etc. It’s just receives less attention in the press because it’s not as sexy as the legislative horse race.


      • ” Rule-making bodies implement public process; it has the exact same requirements for notification, public hearing, etc.”

        But none of that means accountability. A rule-making body can spend six months listening to testimony from the people directly affected by a proposed rule, and then say “nope, we’re still gonna do it like we planned”, and none of those who gave testimony have anything more they can do. That’s the rule now, The End.


      • All systems have trouble learning from mistakes. Markets too. (The tech boom is on its way back, for example.) But lots of governments have done a good job adapting lots of institutions to making a better world.

        But here is the problem. The U.S. system of governance with all its choke points to kill small bore reforms has become the tool of an insanely stupid Republican party that is under the influence of idiot/self-serving-liars like Limbaugh.

        Have Democrats acted badly. Sure, I guess you can find an example. Tu quoque. But the problem is 99% the Republicans and 1% punchable-hippies. This is clear as day.

        Look at the health reform debate. The creation of exchanges and Medicaid expansion to reduce the problem of the uninsured (through community rating and small subsidies on the exchanges, and soshulism) is toxic anathema to the R’s, despite being small bore and mostly conservative.

        They aren’t in support of changes and a system that adapts to reality at all. Not even a little. I guarantee you that if they had given bipartisan support, they could’ve gotten a lot of what they wanted. But they aren’t in favor of bargains and small bore reform that will adapt over time to deal with a changing world.

        Think they’d support Singapore’s system? No way.


      • Market actors do still make mistakes, but market self-correct better because failed experiments in the market will run out of money.

        And I don’t dispute that the Republican party is causing some serious problems, my point is that the aversion to learning in government comes from fundamental incentives of democratic politics, and have little to do with the idiosyncrasies of US political parties.


      • It depends on the market and the governmental program. The healthcare market doesn’t self-correct so good. The market for iPhones, sure.

        We disagree about the causes for sure. The cause is a Republican party who will not accept reasonable small-bore policy changes favored by academics and experts and are now willing to use the choke points in the poorly devised US legislative-executive-judicial system to kill any reasonable reform.


      • It would be a complex discussion to try to tease out how fair this comparison (equivalence?) is, but as a first set of data to consider, I think we can’t go too far astray by reviewing the vote for welfare reform in Congress.

        Democrats voted 98 to 97 in favor of passage with two not voting in the House, and 25 to 21 in favor of passage with one not voting in the Senate. I’m also unaware of a history in which, subsequent to enactment, Democrats then went on to be unwilling to entertain revisions to the law even in the direction they would be traditionally expected to advocate out of a desire to avoid legitimizing the reform ex-post, as Republicans nearly uniformly continue to do more than four years after passage of ACA. Maybe I can enlightened about the existence of that record, however.

        Both sides do things. That’s true. Sometimes those things have similarities. That’s true too. Whether they do the same things understood properly in context (which is necessary for the claim that they both do “it” to be true in a meaningful way), is far less clear. In some cases, it’s probably true. It’s not clear to me that the “it” is a single thing in the case of these two examples, however.


      • …I’d suggest that Bush’s Social Security reform attempt is probably the better comparison. Obviously, there you did run into uniform, intractable opposition that would have taken the same tack in the aftermath had reform been enacted.

        It also highlights what in my view is the important question about critiques of parties for being unwilling to work with majorities in helping “tinker” with reforms enacted pursuant to the agenda of the majority. That is: does the minority party allow that the basic aims of the reform are desirable regardless of its preferences about the policy means and mechanisms being chosen – or even just not grossly negative? (I credit Mark Thompson with helping me to consider this point in the course of discussions during the process of ACA’s passage in 2009-10.) Ultimately, it’s meaningless to critique a party for refusing to play ball on a policy agenda that is fundamentally anathema to the party’s policy commitments.

        So the important question becomes assessing whether the policy program being pursued is truly anathema to the minority’s policy commitments. Social Security privatization clearly was for the Democrats. Republicans claim that ACA was. But Democrats look at Republican’ past policy proposals and see not earnest inability to work with Democrats on ACA out of fundamental policy disagreement, but instead a large degree of politically motivated uncooperativeness (which motivation they also claim was betrayed by various comments by Republicans at the time). Democrats may be wrong to interpret the policy-position history that way, and may be opportunistically seizing on unrepresentative comments to advance that view. It’s open to debate. But this point – whether or how much a policy program is truly fundamentally anathema to a minority party’s policy commitments – is the one that’s most important to consider when assessing their willingness to work with the opposing party to help enact, lend some bipartisan support to, and help refine the majority’s reform efforts. IMO.

        Interestingly, this whole discussion runs in a direction somewhat contrary to the thrust of James’ post, as this kind of bipartisan tinkering tends to produce complex, less transparent legislation, or so I’m guessing. In any case, it helps produce legislation, and legislation produced in a democracy as complex as ours is pretty likely to be complex itself. So if complex legislation is really worse than no legislation as aq semi-axiomatic matter, maybe this tendency toward unwillingness to work across the aisle largely out of a desire to thwart the majority’s perceived effectiveness at governing and thus deal them a political blow is a positive tendency.


    • It’s not obvious to me. As a matter of fact, whenever someones has to point out that something is obvious, it usually means that it is not at all.

      What is obvious to me is that if conservatives didn’t exist, progressives would have to invent them to explain why all their best-laid plans didn’t do all the things they were intended to do.


  10. James K.,

    As others, including , have pointed out above, the issue isn’t so much whether you’re right. It’s whether what you propose is doable. If a political system doesn’t favor the type of policy formation that approaches the standards you outline, then we either have to change the political system, do nothing, or do something. with all the perverse incentives that come along with that.

    Much has been said on this thread about what we would be able to do with a parliamentary system. (What they really mean, by the way, is a system with parliamentary superiority…..we could theoretically have a parliament with institutional checks similar to our congressional system. The Canadian system comes to mind. It does not have as many checks and veto points, but as a federal system, with a Senate and with a tradition of judicial review that actually precedes the 1982 charter of liberties, it has some similar checks.) Do we really want a government with so fewer veto points than the US’s? Maybe (for example the filibuster…..although it looks a lot more attractive when it’s my party in the minority). But if it’s easier to do things right, it’s also easier to do things wrong.


Comments are closed.