I’m going to disagree thoroughly here and say that the root of this problem is not that it is impossible to successfully balance flexibility and accountability, but rather that you are attempting to solve a problem with institutional tools when it really needs to be solved with social ones.
Fundamentally, the problem you have here is that you are neglecting a potential third variable, which is ‘culture’. Patrick Hoverstadt puts this well when he talks about Theory X and Theory Y in employment. Effectively there are two attitudes employees can take in a business - either the attitude that we are all in this together to work towards a positive common goal, or the attitude that I will do what I can to get my money, screw all of you.
So long as you have the first attitude, whatever balance you strike between informality and regulation should hopefully avoid the problems of corruption - whilst it may not be optimal, you can at least deduce the optimal construction of institutional systems. Once you are in the latter however, everyone immediately starts looking out for themselves and you get all the endless problems you describe.
Fundamentally the problem you have here isn’t that there is no perfect balance, but rather that you are attempting to use structural tools to solve problems that their impacts are themselves a function of. You cannot simply regulate or deregulate bad culture away.
That's an interesting point which deserves a longer response - I worry though that the structural responses and move to Theory Y tend to be the results of adversarial outcomes which you can't deal with culturally.
In particular at the national level, the Mont Pelerins and their followers/descendants (or whatever shorthand you prefer for that lot) won and undermined the trust culture and there is no obvious way to get it back. (Certainly at minimum it’s an effort at least as long term and resourced as they undertook.) So we’re stuck working in the culture we have now.
"Rather than optimising a tradeoff, we just react to local conditions and ask questions like 'are the problems we’re currently facing more like the ones we’d expect to see in an excessively rulebound and inflexible system, or in a corrupted and untransparent system?'."
That description points towards another dynamic which will increase oscillation-- people who interact directly with the system will build up an opinion on that question, and then need to communicate to decision makers (management and/or the public) which often involves simplifying and overstating.
There's a market for transforming the opinion, "we have an excessively rulebound and inflexible system" into, "we are facing a crisis."
I think that’s right. Arguably an entirely static pendulum wouldn’t even be desirable given the need for feedback but equally some demarcation of the amplitude of the pendulum’s swing is helpful. I think of it in terms of a machine gun emplacement (bear with me) - you don’t anchor the gun as you need to be able to respond to different targets but equally you stake the traverse of the barrel either side so you don’t swing around so far that you’re gunning down your own troops
I’d add that analog feedback control almost always involves oscillation. It’s only in the digital era that it got easier to lock at a level, but of course human systems just aren’t digital.
I'd think about running policy like a hedge fund. Get a lot of analysts and information and make better bets. You're not seeking one outcome (fnancial return) like the fund, it'd be more like a basket of acceptable returns/outcome ranges. So many people housed, what have you. With AI, even LLMs, this is going to become dramatically easier, and the "too much data/too many factors" dramatically less important. If we don't do this, things may get much worse...a guy can't beat Wall Street, but he could beat the social environment...determine an unfairly advantageous place to live, seen in terms of the free rider problem.
One of the rare instances in which a text might be clarified (rather than obscured) by a discussion of dialectics. The pendulum, as reactive/adaptive, is still better than a static/mechanistic system. But the hope is that the meta system is in fact “learning” something with each oscillation, that might make the swings less erratic—though in practice this will depend on the distribution of shocks to the system, the hope is that the system progresses in H*gel’s “upward spiral of negations”
"Maybe this kind of oscillation between overcorrections is itself the equilibrium. Viability at the system level doesn’t have to be a single steady state. "
Delays in the feedback response can result in stable oscillations, like the famous predator-prey populations. Another possibility is that the forces drive the 2 state stability. They push with increasing force the further away from the central point, reach a maximum, then fall back towards the center, overshoot, as the opposite force then increases. Isn't this something that happens with the centralization vs decentralization in organizations - a constant dynamic that can push one way or the other based on new technology and business fads?
Well, I definitley think there are multiple stages where the pendulum is in a viable solution. I think the problem is more finding what variation is acceptable - i.e. how far the pendulum can go.
The fundamental issue is that politicians - the ones who create bureaucracies and regulations- are in the business of pleasing donors and supporters, not in the business of finding broad-based, sensible, win-win solutions.
I’m going to disagree thoroughly here and say that the root of this problem is not that it is impossible to successfully balance flexibility and accountability, but rather that you are attempting to solve a problem with institutional tools when it really needs to be solved with social ones.
Fundamentally, the problem you have here is that you are neglecting a potential third variable, which is ‘culture’. Patrick Hoverstadt puts this well when he talks about Theory X and Theory Y in employment. Effectively there are two attitudes employees can take in a business - either the attitude that we are all in this together to work towards a positive common goal, or the attitude that I will do what I can to get my money, screw all of you.
So long as you have the first attitude, whatever balance you strike between informality and regulation should hopefully avoid the problems of corruption - whilst it may not be optimal, you can at least deduce the optimal construction of institutional systems. Once you are in the latter however, everyone immediately starts looking out for themselves and you get all the endless problems you describe.
Fundamentally the problem you have here isn’t that there is no perfect balance, but rather that you are attempting to use structural tools to solve problems that their impacts are themselves a function of. You cannot simply regulate or deregulate bad culture away.
That's an interesting point which deserves a longer response - I worry though that the structural responses and move to Theory Y tend to be the results of adversarial outcomes which you can't deal with culturally.
In particular at the national level, the Mont Pelerins and their followers/descendants (or whatever shorthand you prefer for that lot) won and undermined the trust culture and there is no obvious way to get it back. (Certainly at minimum it’s an effort at least as long term and resourced as they undertook.) So we’re stuck working in the culture we have now.
"Rather than optimising a tradeoff, we just react to local conditions and ask questions like 'are the problems we’re currently facing more like the ones we’d expect to see in an excessively rulebound and inflexible system, or in a corrupted and untransparent system?'."
That description points towards another dynamic which will increase oscillation-- people who interact directly with the system will build up an opinion on that question, and then need to communicate to decision makers (management and/or the public) which often involves simplifying and overstating.
There's a market for transforming the opinion, "we have an excessively rulebound and inflexible system" into, "we are facing a crisis."
I think that’s right. Arguably an entirely static pendulum wouldn’t even be desirable given the need for feedback but equally some demarcation of the amplitude of the pendulum’s swing is helpful. I think of it in terms of a machine gun emplacement (bear with me) - you don’t anchor the gun as you need to be able to respond to different targets but equally you stake the traverse of the barrel either side so you don’t swing around so far that you’re gunning down your own troops
You don't need to apologise about that one to me - the problem of over correcting in machine gunfire was the original genesis of cybernetics!
I’d add that analog feedback control almost always involves oscillation. It’s only in the digital era that it got easier to lock at a level, but of course human systems just aren’t digital.
Very nice. The people at 'Protocolized' have this nice discussion of 'tensions', which they describe as 'trade-offs plus conflict', which are to be managed rather than solved, which speaks to some of what you say: https://protocolized.summerofprotocols.com/p/one-tension-to-rule-them-all
I'd think about running policy like a hedge fund. Get a lot of analysts and information and make better bets. You're not seeking one outcome (fnancial return) like the fund, it'd be more like a basket of acceptable returns/outcome ranges. So many people housed, what have you. With AI, even LLMs, this is going to become dramatically easier, and the "too much data/too many factors" dramatically less important. If we don't do this, things may get much worse...a guy can't beat Wall Street, but he could beat the social environment...determine an unfairly advantageous place to live, seen in terms of the free rider problem.
Is there a way to read your "Beyond Accountability" piece if you don't subscribe to Starling Insights?
One of the rare instances in which a text might be clarified (rather than obscured) by a discussion of dialectics. The pendulum, as reactive/adaptive, is still better than a static/mechanistic system. But the hope is that the meta system is in fact “learning” something with each oscillation, that might make the swings less erratic—though in practice this will depend on the distribution of shocks to the system, the hope is that the system progresses in H*gel’s “upward spiral of negations”
"Maybe this kind of oscillation between overcorrections is itself the equilibrium. Viability at the system level doesn’t have to be a single steady state. "
Delays in the feedback response can result in stable oscillations, like the famous predator-prey populations. Another possibility is that the forces drive the 2 state stability. They push with increasing force the further away from the central point, reach a maximum, then fall back towards the center, overshoot, as the opposite force then increases. Isn't this something that happens with the centralization vs decentralization in organizations - a constant dynamic that can push one way or the other based on new technology and business fads?
Well, I definitley think there are multiple stages where the pendulum is in a viable solution. I think the problem is more finding what variation is acceptable - i.e. how far the pendulum can go.
The fundamental issue is that politicians - the ones who create bureaucracies and regulations- are in the business of pleasing donors and supporters, not in the business of finding broad-based, sensible, win-win solutions.