I read Road to Serfdom. I wasn't impressed.
More of Mark's comment:
Mark writes that meddling bureaucrats wish away the negatives of their regulatory mechanisms1. It is not quite that simple. There are very many times when bureaucrats know there will be negative outcomes, but accept them as trade-offs because the overall situation will still be better. There are times when they’re surprised by negative outcomes. There are times when what one group considers a negative outcome is what the bureaucrats considered the point. But I don’t think there are lots of times when bureaucrats ponder a negative outcome, then airily dismiss it. Every negative outcome has a constituency dedicated to making that consequence the most important thing in a bureaucrat’s life. We do not get to dismiss much.
So I don’t think that bureaucrats are willfully making bad regulations, but I have to acknowledge that the accumulated codes are painfully dense thickets, too tangled for easy use by our citizens, complicated, with internal contradictions. I want to talk about how civil servants doing their best end up making systems that are so frustrating that “bureaucrats” and “regulations” are nearly epithets. I mean, Bush’s Executive Order was his usual self-serving corrupt bullshit, but people were willing to give it a pass, just on the grounds that it might make the bureaucrats less able to do the work the country pays them for. How did it get so bad?
Some of it is that we are, like Lindblom said, muddling through. Most of it the rest of it, I think, is that bureaucracies contain programs, which spend their time solving problems. The last of it, I guess, is a lack of vision.
Lindblom pointed out in 1959 that when bureaucracies go to solve problems, they do not evaluate the full range of solutions available, and pick the optimal solution. Instead, they look at a few solutions that look very like the current situation, only a trifle better, and pick one of those. Lindblom says muddling through isn’t all bad, and I agree. But it leads to complexity. It maintains and builds on the existing regulations; no branches get pruned out, but twigs are added for redirection.
So we start with a complex set of regulations, but I think a bigger problem with bureaucratic function is that bureaucracies are divided into programs and work to solve problems. The programs are a good rough cut at pieces of the natural world. On a big scale, we have agencies for air, transportation, forests, agriculture, coasts and oceans, wildlife, water2. Hell, we have two for water. But real problems exist in the continuous system. Salmon live and are fished in the ocean, but swim up streams polluted by agriculture unless they are blocked by road culverts or unless they die in streams that are too warm because too much water was diverted to send to a far away city, or they make it all the way up to headwaters, where their spawning gravel is silted up because the logged hillsides are eroding. Which agency fixes that? Well, they ALL do, piecemeal, applying the lessons of their own fields. You know that regulation will overlap and be inconsistent, because there isn’t a Salmon Agency3 to keep it all in order. It is better for each agency to do something, because any missing piece makes the rest moot, but it will be a jumble when they’re done.
A worse structural flaw, however, is that programs want to solve problems. They are assigned problems by the legislature and the problems are just too narrowly defined. Bureaucrats are pretty good at solving the problem in front of them (creek is flooding? Make it concrete to decrease turbulence and increase flood capacity) and, like all people, not that good at asking how that fits into the world or what effects it will have. They solve the problem in front of them and make more. In some ways it isn’t fair to hold them responsible for the next set of problems (concrete rivers are ugly, no fish or wildlife); they probably did a very nice job solving the problem they were assigned4. But a narrow focus on solving the problem in front of you makes more problems, and agencies have been chasing the problems they’ve caused for decades now. That approach, and its iterations and reversals, has rightfully earned the public’s contempt.
So what is an agency to do if the public demands action about something in its jurisdiction? How can it handle that without adding to already complicated regulation, or making a new problem somewhere else? In the real world, I don’t know. Things probably have to keep going the way they do, muddling through. I have a fantasy though. I imagine sometimes, that instead of figuring out what we gotta do to our regulations to get the results we want, we work backwards from a vision. People who care and people with expert knowledge come up with a detailed picture of the world with that problem solved. Then we ask, what are the physical conditions and the human interactions and the laws that would bring that picture into being? We write those regs, clean and consistent and orderly. That’s the only way I can think to escape the inevitable complexity of an incremental approach.
1 His further comments suggest that he is objecting to the processes we use to develop regulations. EXCELLENT! I wanted to continue this series forever! I am done with funny, pithy posts! Short, lighthearted posts are a thing of the past! We’ll never talk about water or boys or Los Osos again! Who even wants to? Not me!
2 I have heard there are agencies that deal with things that aren’t the environment, but you couldn’t prove that by me.
3 We tried to have a Bay-Delta Agency (called CALFED), but it collapsed. I don’t think that is proof that an inter-field agency organized around the issue is doomed, but the first big one we tried didn’t work.
4 You know what kills me? All those engineers would be just as happy with harder problems. It isn’t that they are attached to the all-concrete river solution; it is just that that is the best solution for the only-flood problem. If you give them a bigger assignment, like flood + nice river-park + fish, they would just as happily optimize that assignment.
Too often, enthusiasts of top-down problem solving seem to address the task of finding a good mechanism by first wishing away all those negatives. There is no benevolent and wise Philosopher King. Many people seem to ignore that fact in their enthusiasm to get to the solving of problems and righting of wrongs. The regulatory mechanism is going to be slow, hidebound, favor those with political clout, and will create unintended side effects as people respond to new incentives.
Mark writes that meddling bureaucrats wish away the negatives of their regulatory mechanisms1. It is not quite that simple. There are very many times when bureaucrats know there will be negative outcomes, but accept them as trade-offs because the overall situation will still be better. There are times when they’re surprised by negative outcomes. There are times when what one group considers a negative outcome is what the bureaucrats considered the point. But I don’t think there are lots of times when bureaucrats ponder a negative outcome, then airily dismiss it. Every negative outcome has a constituency dedicated to making that consequence the most important thing in a bureaucrat’s life. We do not get to dismiss much.
So I don’t think that bureaucrats are willfully making bad regulations, but I have to acknowledge that the accumulated codes are painfully dense thickets, too tangled for easy use by our citizens, complicated, with internal contradictions. I want to talk about how civil servants doing their best end up making systems that are so frustrating that “bureaucrats” and “regulations” are nearly epithets. I mean, Bush’s Executive Order was his usual self-serving corrupt bullshit, but people were willing to give it a pass, just on the grounds that it might make the bureaucrats less able to do the work the country pays them for. How did it get so bad?
Some of it is that we are, like Lindblom said, muddling through. Most of it the rest of it, I think, is that bureaucracies contain programs, which spend their time solving problems. The last of it, I guess, is a lack of vision.
Lindblom pointed out in 1959 that when bureaucracies go to solve problems, they do not evaluate the full range of solutions available, and pick the optimal solution. Instead, they look at a few solutions that look very like the current situation, only a trifle better, and pick one of those. Lindblom says muddling through isn’t all bad, and I agree. But it leads to complexity. It maintains and builds on the existing regulations; no branches get pruned out, but twigs are added for redirection.
So we start with a complex set of regulations, but I think a bigger problem with bureaucratic function is that bureaucracies are divided into programs and work to solve problems. The programs are a good rough cut at pieces of the natural world. On a big scale, we have agencies for air, transportation, forests, agriculture, coasts and oceans, wildlife, water2. Hell, we have two for water. But real problems exist in the continuous system. Salmon live and are fished in the ocean, but swim up streams polluted by agriculture unless they are blocked by road culverts or unless they die in streams that are too warm because too much water was diverted to send to a far away city, or they make it all the way up to headwaters, where their spawning gravel is silted up because the logged hillsides are eroding. Which agency fixes that? Well, they ALL do, piecemeal, applying the lessons of their own fields. You know that regulation will overlap and be inconsistent, because there isn’t a Salmon Agency3 to keep it all in order. It is better for each agency to do something, because any missing piece makes the rest moot, but it will be a jumble when they’re done.
A worse structural flaw, however, is that programs want to solve problems. They are assigned problems by the legislature and the problems are just too narrowly defined. Bureaucrats are pretty good at solving the problem in front of them (creek is flooding? Make it concrete to decrease turbulence and increase flood capacity) and, like all people, not that good at asking how that fits into the world or what effects it will have. They solve the problem in front of them and make more. In some ways it isn’t fair to hold them responsible for the next set of problems (concrete rivers are ugly, no fish or wildlife); they probably did a very nice job solving the problem they were assigned4. But a narrow focus on solving the problem in front of you makes more problems, and agencies have been chasing the problems they’ve caused for decades now. That approach, and its iterations and reversals, has rightfully earned the public’s contempt.
So what is an agency to do if the public demands action about something in its jurisdiction? How can it handle that without adding to already complicated regulation, or making a new problem somewhere else? In the real world, I don’t know. Things probably have to keep going the way they do, muddling through. I have a fantasy though. I imagine sometimes, that instead of figuring out what we gotta do to our regulations to get the results we want, we work backwards from a vision. People who care and people with expert knowledge come up with a detailed picture of the world with that problem solved. Then we ask, what are the physical conditions and the human interactions and the laws that would bring that picture into being? We write those regs, clean and consistent and orderly. That’s the only way I can think to escape the inevitable complexity of an incremental approach.
1 His further comments suggest that he is objecting to the processes we use to develop regulations. EXCELLENT! I wanted to continue this series forever! I am done with funny, pithy posts! Short, lighthearted posts are a thing of the past! We’ll never talk about water or boys or Los Osos again! Who even wants to? Not me!
2 I have heard there are agencies that deal with things that aren’t the environment, but you couldn’t prove that by me.
3 We tried to have a Bay-Delta Agency (called CALFED), but it collapsed. I don’t think that is proof that an inter-field agency organized around the issue is doomed, but the first big one we tried didn’t work.
4 You know what kills me? All those engineers would be just as happy with harder problems. It isn’t that they are attached to the all-concrete river solution; it is just that that is the best solution for the only-flood problem. If you give them a bigger assignment, like flood + nice river-park + fish, they would just as happily optimize that assignment.
posted by Megan at 12:36 AM
15 Comments:
I believe this carries us nicely into the next phase of the series, on public decision-making processes, but if I left anything about bureaucracies, bureaucrats or regulations, you go ahead and tell me.
Don't look for me this weekend, but we'll continue this streak next week. Next week and forever, until you beg for mercy.
"There are very many times when bureaucrats know there will be negative outcomes, but accept them as trade-offs because the overall situation will still be better."
Better in whose opinion? These agencies aren't part of the political process. Would it be ok for one of these agencies to create a new regulation that is wildly unpopular, but in their opinion for the best?
I don't really buy the benevolence of the people working at these agencies either. Everyone has their own politics, and beliefs, so if people with a particular political persuasion are more likely to work for a certain agency, then their regulations will reflect that, and there's nothing the rest of us can do about it.
"Every negative outcome has a constituency dedicated to making that consequence the most important thing in a bureaucrat’s life. We do not get to dismiss much."
That's hardly comforting either. So, you're appeasing every special interest? It seems to me that would bias you towards over regulation, rather than towards a moderate amount.
", just on the grounds that it might make the bureaucrats less able to do the work the country pays them for."
You understand that for a lot of us the problem is that we don't want you doing much of this work. For a lot of us, we'd like to see the size of the government shrink, and the amount of regulation reduced.
In general, there's no reason the laws of this country should be so complicated. There's something wrong when the average person never has any idea whether or not he's doing something illegal.
Every big engineering project I've ever seen comes with this same kind of incremental buildup of complexity that just doesn't need to be there, but at the time it's expedient. Of course, every big engineering project eventually ends. You get the chance to scrap as much of that infrastructure as you want and start from scratch.
Governments rarely do that. They just keep growing. Once a law is on the books it's hard to get rid of it.
The same goes for these government agencies. Once they're created they rarely go away. But they do tend to grow.
The original point was elected officials should be more accountable for these agencies at least. In fact, I'm still in agreement with the link I gave. All of these regulations should have to go before Congress, and through the process like any other law before taking effect.
You have a lot of faith in these agencies, and the people working in them. I don't, and a lot of people agree with me. I don't want someone doing what's in my best interest for me, I'd much rather make those decisions for myself, and at the very least have the opportunity to vote for the person representing me.
I hope it is not against your policy to post links but this is very interesting:
http://www.newamerica.net/publications/articles/2005/in_defence_of_mandarins
if that did not come out properly, it is an online article by Michael Lind in the New American prospect called 'In Defence of Mandarins'.
here's Billo's link.
What you gave was broken Billo, I found it with google.
Please stop. I miss cool Megan. I do not miss mandarin Megan. -K.
"Road to Surfdom", is that about PCH or something?
"The power which the gnarliest millionaire dude, who may be my bro and perhaps my employer, has over me is totally less than that which a weak sauce functionaire possesses who wields the bogus coercive power of the state, and on whose discretion it depends whether and how I am able to shred."
megan, i think your post is insightful. i have the same problem at work when i write code. everyone wants to solve the latest little problem and they never look the bigger picture and they never take the time to refactor the code. instead we end up with this horrendous spaghetti code which is complicated, buggy, difficult to understand, and difficult to add features to.
i really like your explanation for the complexity. i have never heard it before.
i don't think you really proposed a solution and i dont have one either. not even at work. i know smart people are genuinely trying, but somehow the incentives just are not aligned correctly. i guess i think i should be ceo.. but beyond that...? i have no idea.
i gotta agree with justin and say this is a reason to limit government as much as possible. but clearly we do need some government. is endless spaghetti code the best of a bad set of scenarios?
I believe government and regulations are necessary but overused. It's a shame we can't just pick up and start a new colony on Mars, using just the bare minimum of rules that are actually needed. Oh, in a hundred years or so you'd have to leave because things would have gotten just as badly tangled by then, but that way the ones who wanted the nanny could stay, and the ones who wanted the freedom could go....
We miscommunicated mildly, but in an important detail. I am not objecting so much to the methods and processes used by the regulators and bureaucrats as I am the STRATEGIC decision to throw the problem their way to begin with.
We seem to be in agreement that the regulatory process is messy and imperfect and has a lot of negative aspects. And I've already granted that in some cases, it's probably the best among all flawed options.
But the following chain of "reasoning" is disturbingly ubiquitous:
1) The current process for handling situation X has flaws
2) Therefore, let's make a government agency responsible for fixing the problem.
3) If we can just get all the evil people to shut up and get out of the way, we can make everything perfect.
For a good chunk of advocates, opinion-makers, and voters, any proposed top-down regulation scheme is viewed in the light of maximal optimism, whereas a price-mechanism or individual-choice mechanism is always viewed with maximal suspicion.
And I feel compelled to reiterate that some issues, like water management, are probably best handled by throwing them into the sphere of public regulation.
Others, like taxi-cabs and education, seem much more debatable.
Isn't it serfdom? Or am I missing the joke?
Dudes, you're so right. That's what I get for posting late at night. I'm gonna fix it, but the comments can memorialize my error forever.
Megan et al.: Great discussion and comments. Bureaucrats are not so "efficient" for two reasons:
1) They are not us, but they make decisions for us. They WILL get those wrong for at least somebody.
2) They have persistent market power. By this, I mean they do not go out of business, and you don't get to choose which of their great ideas to adopt.
Programs are troubling for another reason -- they are inflexible. At the time many programs are created, they are the OUTCOME of a complex negotiating process among interest groups. After that point, however, things change, but the programs do not. The "by the book" bureaucrats enforce the law/codes/etc on a population that suffers more and more as circumstances change.
Contrast these characteristics with a semi-competitive market: businesses compete for customers, who can go where they wish. Businesses will change their operations to increase profits on an ad-hoc basis -- or be driven out of business by others that do so better.
A GREAT book on this topic is Bureaucracy: What Government Agencies Do and Why They Do It by James Q. Wilson
ps/Another aspect of Bureaucracy that Hayek appreciated and I hope you do too is that bureaucrats qua people cannot understand how everything connects. Nobody can, of course, but bureaucrats impose their understanding on us. I am GREAT at understanding what works for me, and I can change what I do as I learn from my mistakes (oops -- don't like quiche), but bureaucrats have little or no positive/negative feedback from us. Hence Hayek's good idea: let people run their own lives, with clear property rights that facilitate voluntary interaction...
1) They are not us, but they make decisions for us. They WILL get those wrong for at least somebody.
The standard isn't to make decisions that will make things right be everyone. The standard is to make decisions that will make more things right than were before. Keep in mind that the No Action Alternative is often considerably flawed.
but bureaucrats impose their understanding on us... I am GREAT at understanding what works for me... but bureaucrats have little or no positive/negative feedback from us.
But, it isn't all about you. It doesn't always work well, but what a bureaucrat is doing is using professional judgment (years of expertise in whatever) to make regulations that balance competing interests in a flawed but pre-defined process.
Who would you rather have do that? Yourself? You can't have enough expertise in enough fields and meet enough people to internalize your externalities or get remedies from the people who accidently injure you. The Queen? She can't be expert enough in enough fields to manage them all. Legislators? They give us broad direction and we do the technical work.
You guys are so quick to identify with the person who gets constrained by regulation. There are many, many interests in our complicated society; just because you got the shaft on some regulation doesn't mean that it wasn't a better move overall. You are the beneficiary of hundreds of regulations you don't even know exist.
I talked about feedback. Show up at my meetings and make my regulations better.
I think we should not underestimate the importance of professionalism.
We take it as a given (I hope) that military officers and priests obey a higher calling.
Yet we know the Catholic Church and the Pentagon are horrendous, self serving bureaucracies.
Yet we trust one to defend our country (in my case the British Army and its regiments) and the other to mind our souls (scandals notwithstanding).
The reality is many people go into public service, or a profession, out of a sense of calling.
What they do is an integral part of what they are.
People are much happier being lawyers or surgeons on $500k pa, than being real estate salesman on $500k pa. There is an implicit valuation of a professional calling in that.
This doesn't make them perfect decision makers, but it is probably a key factor in understanding how and when professional organisations in the public service (be it the Fire Department of New York, the US Marines Corps, or a good public school) operate effectively.
I can wander round the cities of the United States and tell you that some, demonstrably, work better than others.
I suspect the professionalism of the much despised bureaucrats has something to do with that.
Valuethinker
Post a Comment
<< Home