The Well Intentioned Commissar

http://highered.blogspot.com/2009/01/well-intentioned-commissar.html

by ColinWright • 14 years ago

75 29 14 years ago

highered.blogspot.com

wpietri• 14 years ago

If this has some resonance for you, I'd encourage you to check out one of the Lean Manufacturing books. The Toyota Way and Toyota Kata are two of my favorites. Also great is the "This American Life" story NUMMI: http://www.thisamericanlife.org/radio-archives/episode/403/n...

I bring up Lean Manufacturing because they have a strong analytical approach that looks at all operations in terms of whether or not they create customer value. Studying the various Lean techniques (e.g., value stream mapping, one-piece flow, kanbans) has given me a lot of insight into ways that typical business thinking goes wrong.

Lean Manufacturing to me is something that happened when engineers maintained control of a business.

Game_Ender• 14 years ago

Thanks a lot for sharing that NPR story. It was very enlightening to learn more about the differences between the GM and the Toyota production methodologies.

Zarkonnen• 14 years ago

For a very good extended historical view on how this kind of mess-up kept on happening in communism, also see the book Red Plenty. ( http://www.amazon.co.uk/Red-Plenty-Francis-Spufford/dp/05712... )

zeteo• 14 years ago

If you have any interest at all in the subject, Zarkonnen is making a very good recommendation. Red Plenty has brought Soviet society to life for me in a way that no other book or article could (and I've read a few).

temphn• 14 years ago

Reminiscent of Eliyahu Goldratt's excellent book, The Goal. Bottom line is that unless you incentivize on profit, you are incentivizing loss.

greenyoda• 14 years ago

Don't miss the punch line, which is at the very bottom: "Permission to copy: You may share copies of this story with anyone who might enjoy thinking about its implications in academia."

So this article is really about churning out useless research in academia, where merit is measured by how many publications you have rather than by what actual benefit the research has.

farnsworth• 14 years ago

Or by trying to use measures like student evaluations, avg. grade, etc. as a measure of professor quality. It's always awkward to apply a bureaucracy to something as 'pure' as education and research.

davewasmer• 14 years ago

The real moral of the story: prices are the only effective mechanism for the efficient allocation of higher order goods.

nn2• 14 years ago

Except when they don't work, like in healthcare, lawyers, research etc.

relevant recent HN stories: healthcare: http://news.ycombinator.com/item?id=3661731 lawyers: http://news.ycombinator.com/item?id=3672815

apenwarr• 14 years ago

Many, many people have misinterpreted this kind of story and tried to apply prices to everything (eg. "buying" support time from staff in the IT department). That doesn't work either, because you can do the same tricks with money as you can with any other metric.

For example, in the dotcom days, lots of companies would inflate their revenues (as much as they wanted!) by simply buying ads from each other, back and forth.

DanI-S• 14 years ago

This is certainly a wise parable. However, the way in which we routinely use stories of this nature from Communist countries as a source of kitsch hilarity leaves a bad taste in my mouth.

Mismanagement and the subversion of metrics to give the illusion of success happen universally, whatever economic system you live under. To some extent, it's human nature to game the system. By using these cliched examples, we're intimating that our society is somehow different.

In reality, our atrocious metrics for success have caused us to spend the past couple of decades slinging fantastic financial rewards at those who have inadvertently sabotaged our economy in their hunt for ever-greater returns.

spindritf2• 14 years ago

> By using these cliched examples, we're intimating that our society is somehow different.

It is, the price system is fundamentally different. Even if those or similar processes take place within certain organizations (like large corporations or universities), most absurdities are largely confined to those organizations.

And the examples from communist countries are great because they show where bureaucratically controlled economy leads, which is sometimes difficult to notice in systems where government's influence is more or less diluted.

theaeolist• 14 years ago

Yes, these absurdities only happen in large corporations, universities, hospitals, governments and other large organisations. Even though hot-dog stands and corner stores are free of such aberrations I'd still worry.

DanBC• 14 years ago

I'm uncomfortable with naming a nation. Especially with the US cultural thing of "polish jokes". But the article needs a nonsensical bureaucratic system, and CCCP communism provides an excellent example.

I agree with you that a poor metric will do harm, and that there are plenty of examples in "western" manufacturing. (I used to work in electronic sub contract manufacturing, and I saw the nonsense that was ISO900x documentation.)

Going further, there are a few things which need new names.

"Mexican Bridge" - there's a four lane road bridge and politicians want to improve traffic flow so they remove the lines and paint more, turning it into a 6 lane bridge. That's a 50% increase in lanes, but now they are narrower and so there are more accidents. The politicians decide to go back to four lanes. That's a 33% reduction. Because they are politicians they call this a 50% - 33% = 17% increase. (This is from "How to lie with statistics", a great book with occasional uncomfortable language.)

"Chinese Whispers" - the childhood game where one child starts by whispering a sentence to her neighbour, who whispers it to his neighbour, etc, and then at the end of the chain you compare the original with the final sentence. I've heard this called "Broken Telephone"; but no-one (in the UK) knows what I mean when I call it that.

"Indian Giver" Louie CK does an excellent routine on this.

And then there are other things which are fine, but which some people think aren't: "Nitty Gritty" (nothing to do with slave trading ships) "Chinese wall" (in business, no racist connotation) etc.

spindritf2• 14 years ago

> I'm uncomfortable with naming a nation.

This is the sort of a true lie ("prawdziwe zmyślenie") that accurately describes the situation in the People's Republic of Poland, no need to pretend otherwise.

nirvana2• 14 years ago

"In reality, our atrocious metrics for success have caused us to spend the past couple of decades slinging fantastic financial rewards at those who have inadvertently sabotaged our economy in their hunt for ever-greater returns."

This is exactly true, but it is also why these parables are relevant. The "atrocious metrics" you refer to were not created by the free market, but by the very same people who created them in the parable-- commissars.

For instance, in 2001-2002, after the dotcom crash, I knew there was going to be a housing boom because the federal reserve made the cost of money lower than the rate of monetary inflation. (I didn't figure this out myself, though, but learned it from some wise economists.) Banks were incentivized to lend as much as possible, because a housing boom is what the commissars wanted (Paul Krugman wrote about it, for instance.) This created a lot of jobs and while it also created the real estate bubble and consequential crash, it was good for the Bush administration.

The interest rates, the regulations on banks, and the inflation rate are all metrics created by the government. The low rates and high inflation are metrics that created a carry trade. The Clinton era changes in regulations[1] incentivized loaning money to people who could not repay it. Of course the FDIC, Fannie Mae and Freddie Mac are all government entities that distorted the market and created moral hazard.

Simply eliminating the FDIC (but requiring banks to have deposit insurance) would have mitigated a lot of damage- the banks who were way over levered would have seen their deposit insurance rates go up, and thus be incentivized to not be over levered.

It is not that our economy is immune to the impact of commissars. Our economy is ruled by commissars... we just pretend it isn't.

If we want to eliminate arbitrary formulas like the ones in the parable, balancing the federal government (removing much of the need for inflation) and letting interest rates float in the market (rather than be controlled by the privately owned Federal Reserve corporation) would help.

However, just like the situation in Communist era Poland in the parable, there are major political movements who think this is crazy talk and would like to spin these economic failings as proof that we need more commissars, not less. I can point to many members of both political parties in the USA who feel this way.

[1] Early in the Clinton era there was a lot of muckraking about how banks were not loaning to ethnic minorities. This produced changes in the regulations. Later an independant study was done that controlled for the ability of the applicants to repay and found no bias. Bank weren't "greedily" refusing to loan to minorities, they were greedily loaning to everyone who could repay. The regulations, however, incentivized loaning even to people who couldn't repay, hence the boom & bust.

barrkel• 14 years ago

"the banks who were way over levered would have seen their deposit insurance rates go up, and thus be incentivized to not be over levered."

That relies on the people selling insurance to be rational. Unfortunately, markets can act irrationally for long enough that disaster still strikes when they come to their senses. For example, it took way too long for CDS rates (i.e. insurance on loans) to rise on Greek debt. In the 19th century - before the creation of the Fed - banking crises and depressions were a dime a dozen in the US: http://en.wikipedia.org/wiki/List_of_banking_crises#19th_cen...

Our economy is ruled by people, like our governments are run by people. The higher up you go, the more scope there is for mistakes, and IMO the less likely you'll see good quality market mechanics. Instead, you see power games, financial hostage taking, threats and counter-threats, etc.

Hoff• 14 years ago

What's your goal with that?

Eliminating regulatory agencies and regulations? The repeal of Glass-Stengal had some repercussions, after all.

Or are you looking to lodge blame with one or the other of the political parties? (There's more than enough muck to rake, of course.)

Or might you be interested in looking at the data, and at details such as the "robo-signing" mess and the questionable legality of the MERS assignments underneath the mortgage securitization (with various mortgage transfers which don't look to have been legally registered) and at the data around the melt-down?

And with the recent proposed banking settlement that's underway between various Attorneys General and financial entities, the behaviors probably would have landed any individual in the dock on fraud charges could well turn into a reasonable investment and a write-off for some entities, too.

There's more than enough of a mess here. Look at the data. And look at what folks are telling you, and at their motivations for telling you want they're positing. (And while you're at it, have a look at the wealth distribution in the US, too, and what the trends look like for that.)

Follow the money, in simplest terms. And that doesn't point to CRA.

Some reading:

http://www.ritholtz.com/blog/2010/05/rewriting-the-causes-of...

http://motherjones.com/mojo/2010/05/dear-gop-fannie-mae-fred...

trevelyan• 14 years ago

@nirvana,

I'm curious what specific regulation you believe is responsible. Subprime loans are by definition those which do not qualify for federal guarantees. Banks never had a financial incentive from the government to make subprime loans since they would have been unable to offload those loans to Fannie Mae.

The problem was risk-assessment on the market-level (purchasers of securitized products who were willfully blind or were misled about the levels of risk involved), mortgage providers who were under pressure to create mortgages at scale, and banks which profited from being in a middle-man position bundling and selling mortgage-based products onto private markets.

funkah• 14 years ago

I totally agree. Large, dumb firms in capitalist economies have many of the same bureaucratic inefficiencies and silly management theories. We tell ourselves that they'll suffer for it, but in reality they can become too big to defeat. Comcast and basically the whole US wireless telcom industry are good examples.

rafski• 14 years ago

The story is not really a parable, this is exactly what Communist economy looked like in Poland and other Eastern Bloc countries. Production levels weren't supposed to meet market demand, they were supposed to meet The Plan. As a result, there were shortages of virtually everything but—who would've guessed—vinegar.

Oddly enough Poland revisited central planning idiocy again when it joined the EU with its economic absurdities, e.g. production quotas not allowing farmers to produce more milk or sugar than the plan allows.

A friend of mine living in Poland now told me about a farmer client of hers who visited her company last year with a large truck and a trailer full of cucumbers to give away to employees. The cucumbers were either oddly-shaped or wrong size according to EU norms but perfectly good otherwise. He couldn't legally sell them, was supposed to dump them instead. Not being OK with wasting food he drove around the city to people he knew, trying to give away literally tons of cucumbers...

cheatercheater• 14 years ago

OK, let's amuse your suggestion for a second:

1. I go to a shop, want to buy a cucumber, what I find is oddly shaped with bulges on it. Perfectly edible, but not that great for cooking and therefore I get less value. It is however sold as a cucumber; I am dissatisfied. I understand it might be a good idea to be able to sell those cucumbers; I do not trust your assertion that the cucumbers could not be sold legally - they probably could be sold at prices lower than usual. They could also be sold to companies that turn them into pulp or something like that; as animal food; etc. They just couldn't be sold as the Standard Cucumber Type #23.

2. Farmers are allowed to sell as much as they want. Some of them go for it and are able to sell the extra and end up making more money; others produce a lot of shit and end up not selling. I remember reading the news about road blocks set up by farmers dumping their grain onto highways. Yeah, nice thinking, Commisar Rafski.

Countries which do not have quotas end up with cheap butter. Why? Because the government ends up being forced to buy the surplus milk at standard price. Lots of farmers end up working on the assumption that as much milk as they produce is good, because it'll get bought at that specific minimal price no matter what quality it is. They could instead spend their time and resources making something that's needed, but no, they end up making volatile produce that'll decay in a week if it doesn't get turned to butter. And so, you end up having lots of butter. Again you are incentivizing completely pointless work. Really well intentioned, too. But short-sighted. Think twice the next time you compare the finance and economic prowess of the leading scholars in the whole European Union combined, with that of your farmer friend-of-a-friend who hasn't finished high school (that's the typical educational level of farmers in Poland).

kabdib• 14 years ago

We received a bunch of code from another group in this Large Software Company. More than half the code consisted of extra do-nothing layers of abstraction; functions that merely did checking and assertions on their arguments and passed them down to lower layers. There were many redundant assertions (e.g., checking that pointers weren't NULL, many times, and that assignments had "made it"). Working in this code was awful.

One day it dawned on me: This other group had a requirement that all code pass a seventy percent code coverage bar before checkin. This extra "safe" stuff was padding to make it easier for them to pass.

We took this code, ripped it to shreds and shrunk it by a factor of three or four. Sped it up a lot, too.

I don't believe in blind use of software metrics as a method for improving software.

gathly• 14 years ago

This story took a long time to explain a simple concept.

nirvana2• 14 years ago

I was in Berlin in the fall of 2010 for several months and we went to visit the DDR museum. (This is the museum about the former East Germany, which was the communist controlled half of Berlin and Germany.)

One of the things they had in the museum was a simulation of exactly this situation. You got to be the commissar of an industry and you had to centrally manage the economy. You almost always failed completely. (You had to deal with more than production, though, you had to deal with worker housing, and stuff like that.)

It was very educational. I'd say that at least the curators of the DDR museum agree with this parable being relevant.

abecedarius• 14 years ago

tldr: an illustration of http://en.wikipedia.org/wiki/Goodhart%27s_law

shalmanese• 14 years ago

I'm sick of reading these parables from the perspective of an all-knowing outsider with perfect information. All it does is give the reader a smug sense of satisfaction that they could never be so stupid.

For once, I'd like to see it from the perspective of someone with uncertain, incomplete information, gathered by people who have their own agendas about a topic whose towers of abstraction the protagonist barely has a grasp on, on which they have to make decisions with inadequate time for consideration or research and of which, 999 will be utterly uneventful and 1 will blow up completely in their face.

lazugod• 14 years ago

This parable is trying to demonstrate a point; padding it out with "uncertain, incomplete information" wouldn't make that point clearer.

dbuxton• 14 years ago

I have lots of friends who are teachers here in the UK. The only metric that they really care about is whether or not their pupils achieve a grade C or higher at GCSE (age 16).

A very high level of effort is focused on D/C borderline students as a result, with more or less able students more or less abandoned. The ROI from moving a D to a C is [perceived as] far greater than moving a C to a B.

So much for actual education...

The Well Intentioned Commissar

Related Stories