‘Decisions’ Category

The Out­rage Makes Them Larger

Recently, much has been writ­ten about “Wall Street” bonuses. Almost all of those arti­cles men­tion the same two things: (1) pop­ulist and gov­ern­ment sen­ti­ment against the bonuses, and (2) the com­po­si­tion of the bonuses towards long-​term, restricted stock and away from cash. At least some of the drive towards a more stock-​heavy com­po­si­tion seems to be management’s attempt to appease the gov­ern­ment and the pub­lic. In this post, we argue that such moves are need­lessly costly, which means inef­fi­cient and larger than need be.¹

In a pre­vi­ous post, Gov­ern­ment Whin­ing and Bailout Fees, we dis­cussed the out­rage and men­tioned that cit­i­zens have a right to be angry – at the gov­ern­ment. In this post, we ana­lyze the reported com­po­si­tion of many of bonuses. In par­tic­u­lar, we think the insis­tence on long-​term, restricted stock grants is inef­fi­cient for sev­eral rea­sons that we dis­cuss below.

How­ever, before con­tin­u­ing, it is worth re-​mentioning that much of the con­tro­versy could be elim­i­nated by elim­i­nat­ing pro­pri­etary trad­ing at insured insti­tu­tions. As we have repeat­edly writ­ten, we have noth­ing against pro­pri­etary trad­ing or traders, but see no rea­son why we or other tax-​payers should sub­si­dize trad­ing losses. Note, too, that there are other good rea­sons to elim­i­nate such activ­i­ties at insured insti­tu­tions, includ­ing the fact that they diverts man­age­r­ial atten­tion away from (bor­ing and mun­dane) every­day activ­i­ties of run­ning com­mer­cial banks. We know that at the mar­gin, there’s not much of a dif­fer­ence between a bank’s trea­sury (asset-​liability) man­age­ment and cer­tain kinds of prop trad­ing, but we’d pre­fer that reg­u­la­tors keep a nar­rower focus. Finally, to get, in a sin­gle edi­tion of The Wall Street Jour­nal, Thomas Frank, Jonathan Macey, and James B. Stewart to agree with us is mind-​boggling. It indi­cates the abject per­ver­sity of the sta­tus quo.

Now, hav­ing said that, we hope that every­one receiv­ing the much-​discussed bonuses get max­i­mum enjoy­ment and sat­is­fac­tion from them. We cer­tainly don’t blame any­one for try­ing to max­i­mum his or her com­pen­sa­tion in an attempt to max­i­mize their sat­is­fac­tion, their family’s sat­is­fac­tion and well-​being, and their con­tri­bu­tion to the less for­tu­nate. The prob­lem is that there are likely cheaper ways to pro­vide the same level of sat­is­fac­tion and reward.

Aside: note that for the remain­der of this post, we’ll use the word “expected,” as in “expected com­pen­sa­tion,” in a very loose, non-​mathematical way. That’s because we are rather pedan­tic and like to empha­size the dif­fer­ence between uncer­tainty and risk. Like oth­ers, we define risk as mea­sur­able uncer­tainty, and that means that risk is a spe­cial type of uncer­tainty or unknow­ing can be (appro­pri­ately) described as a prob­a­bil­ity dis­tri­b­u­tion. Not all prob­a­bil­ity dis­tri­b­u­tions have means or expected val­ues, and that seems to be the case in finan­cial mar­kets. So, try­ing to cal­cu­late one’s expected bonus as a func­tion of mar­ket per­for­mance might not be tech­ni­cally fea­si­ble if the dis­tri­b­u­tion of returns is unknown or its moments don’t exist.²

So what’s wrong with bonuses in the form of long-​term, restricted stock?

Well, they are long-​term so they defer con­sump­tion, they are restricted so they’re are expen­sive to con­vert into con­sump­tion, and they in sotck so they are risky (uncer­tain) because they are only very weakly tied to an individual’s performance.

Delayed Grat­i­fi­ca­tion:

Are there good rea­sons for long-​term com­pen­sa­tion schemes? Yes, there are.

When employ­ees take actions or make deci­sions that have long-​term impli­ca­tions, then sig­nals from mul­ti­ple peri­ods can be used to infer whether the employee behaved appro­pri­ately – back when the the deci­sion was made.

Gen­er­ally, the use of mul­ti­ple sig­nals improves the pre­ci­sion of the infer­ence, and that means that less risk is imposed on the employee.³ For risk-​averse employ­ees, that means a lower risk pre­mium is required to ensure his or her par­tic­i­pa­tion, which means a smaller expected bonus is required.⁴ So, the key to reward­ing long-​term per­for­mance is clas­si­fy­ing cur­rent period results into the time peri­ods when deci­sions were made so that one can make bet­ter infer­ences about the deci­sions made in a prior period. It’s not as easy as it sound, but it is pos­si­ble to do.

So, yes, most traders that make long-​term bets should be rewarded on long-​term per­for­mance, and fea­tures like claw backs should be used, but in the spe­cific way that we wrote about in Claw­backs: the Good, the Bad, and the Ugly and Incen­tives at UBS and in Gen­eral.

How­ever, requir­ing some­one to wait five years to receive stock in a mega-​corporation is not the same thing. That’s because:

1. Five years is arbi­trary, and may have lit­tle to do with the length of the employee’s invest­ment deci­sion. More­over, it is a long-​time to wait for a pay-​off.
2. If we’ve learned noth­ing else dur­ing the past few years, we have learned that, in gen­eral, share prices are very volatile, which means that employ­ees who must wait five years for their reward must bear a sub­stan­tial amount of risk.
3. Other than pos­si­bly a few senior exec­u­tives, no sin­gle employee has very much antic­i­pated or expected influ­ence on share price in five years. Ex post they may have, but not ex ante.

So, it seems rea­son­able to con­clude that impa­tient, risk-​averse employ­ees would sub­stan­tially dis­count the expected value of such stock grants.⁵ That means that all things equal, it means that if they can, employ­ees will demand larger bonus grants to com­pen­sate for the delayed grat­i­fi­ca­tion and the risk.

Restric­tive:

We imag­ine that the only peo­ple who pre­fer that bonuses be in the form of restricted stock are folks who aren’t get­ting them and the envi­ous types: please see The Chil­dren who Have Eaten their Cake…

Usu­ally, there are ways to bor­row against such grants and/​or hedge the value of such grants, but not all firms per­mit such actions. More­over, they’re not cheap and they can be time-​consuming.

That means that employ­ees will bear costs of con­vert­ing the awards to nearer-​term con­sump­tion and, if pos­si­ble, will demand larger bonuses to cover those costs.

Risky and Uninformative:

For some reason,many folks (and politi­cians) believe that when employ­ees own shares, includ­ing restricted stock, incen­tives are some­how mag­i­cally aligned – kind of like Lucky Charms.

How­ever, except for pos­si­bly a small hand­ful of very senior man­agers, that’s very silly. Con­sider that Bank of Amer­ica has nearly 300,000 employ­ees, Cit­i­Group has about the same, and even smaller firms like Gold­man Sachs have more than 30,000. So, the effect of any sin­gle employee is usu­ally very small. (More­over, the pre­dicted effect is usu­ally very small. In fact, when it is large, it is often due to the firm’s fran­chise and rep­u­ta­tion and not that par­tic­u­lar person’s actions.)

Do note that attempt­ing to link the effects of a par­tic­u­lar action, deci­sion, invest­ment or trade to share price today or any point in the future is extremely dif­fi­cult. (Maybe not in finance class, but it is in real life.)

Just as impor­tantly, and as we men­tioned above, even if it can be done (in expec­ta­tion) the firm’s stock price is a par­tic­u­larly noisy mea­sure of a par­tic­u­larly person’s per­for­mance. So, it’s quite pos­si­ble to con­clude that employ­ees will ignore the impli­ca­tion of their deci­sion of share prices, which is com­pletely ratio­nal, and do what’s best for them­selves. That very much reminds us of that quote of Huck­le­berry Finn that we always used when we taught: “Well, then, says I, what’s the use you learn­ing to do right when it’s trou­ble­some to do right and ain’t no trou­ble to do wrong, and the wages is just the same?”

For more on this gen­eral topic, we refer inter­ested read­ers to our essay in the Fal­lacies sec­tion of the web site: One Per­for­mance Mea­sure to Rule Them All.

For more on this topic as it per­tains to trad­ing, we encour­age vis­i­tors to read the last half of the above-​mentioned, The Chil­dren who Have Eaten their Cake…

In sum, we argue that (1) the long-​term nature that delays con­sump­tion, (2) the restricted nature that is costly to bypass, and (3) risky nature fur­ther reduces the value (think in terms of expected util­ity or cer­tainty equiv­a­lent) make such bonuses worth sub­stan­tially less than their face value. If employ­ees have any bar­gain­ing or nego­ti­at­ing power, firms will have to increase the stated value of the bonuses to sat­isfy them.

Those extra costs would be worth bear­ing if they aligned incen­tives, but unless you, dear reader, believes in magic, there is no rea­son to believe that any future actions by those employ­ees will be coöper­a­tive in nature.

So, it seems that long-​term, restricted stock awards are inef­fi­cient ways to moti­vate employees.

We’ll likely proof­read this post and edit it in the near future.

P.S. Our New Year’s res­o­lu­tion is to write more about finan­cial mat­ters, the indus­try and the cri­sis than we did dur­ing last half of 2009. Last fall’s drought occurred for a vari­ety of good rea­sons, but two related ones are worth men­tion­ing: (1) while many of our posts tend to be long, we hate being repet­i­tive, and in our mind there was lit­tle new to say, and (2) with lit­tle new to say, we found many of the events and pro­ceed­ing to be quite bor­ing. For writ­ing blog posts, “bor­ing” means too many ref­er­ences to old mate­r­ial – like above – but we’ll try to write more in 2010.

Copy­right © 2010 Spero Consulting

---

Foot­notes:

We really like Ambas­sador John Bolton’s opin­ion col­umn, Let’s Take Bureau­cracy Out of Intel­li­gence, which appears in today’s edi­tion of The Wall Street Jour­nal, and we like it for the usual rea­son: his iden­ti­fi­ca­tion of the cause of government’s recent intel­li­gence fail­ure and rec­om­men­da­tions are sim­i­lar to ours.

That’s what we wrote about last week where we blamed fail­ures on overly-​centralized and overly-​rigid infor­ma­tion systems:

1. Sad but True: Intel­li­gence Fail­ures & Bad Infor­ma­tion Systems
2. Human Error (ver­sus Sys­temic Failure)
3. Intel­li­gence Fail­ures and Bad Infor­ma­tion Sys­tem Design

As he wrote:

The prob­lem is often not the intel­li­gence we col­lect, but assess­ing its implications.

In other words, the data were there, but no one noticed, or some­one did notice, they either: (1) didn’t bother to men­tion it or (2) men­tioned it and no one lis­tened. That’s the prob­lem with overly-​rigid and overly-​centralized infor­ma­tion systems.

In many orga­ni­za­tions, there’s a vital use for large, cen­tral­ized, rigid sys­tems, but that use involves record-​keeping of easily-​understood trans­ac­tions and events and not the stor­age of assess­ments and inter­pre­ta­tions – like whether some­one might be a terrorist.

In that regard, maybe a good heuris­tic for intel­li­gence sys­tem design­ers (and all design­ers) is: if a field or record requires many assump­tions to inter­pret, its place­ment in a large-​scale sys­tem may be sub-​optimal (and data and infor­ma­tion should be retained and ana­lyzed at a local level). We’ll write more about that on another day, but it is the con­for­mance process that doesn’t sep­a­rate the “wheat from the chaff” so-​to-​speak. In other words, the (false) demand for uni­for­mity and stan­dard­iza­tion strips away the sig­nal as well as the noise. For exam­ple, seethe dis­cus­sion about “Rrea­son­able sus­pi­cion” in Human Error (ver­sus Sys­temic Fail­ure).

That’s sim­i­lar to what he wrote:

Each intel­li­gence agency should be able to place its analy­sis of data into a com­pet­i­tive mar­ket­place of clas­si­fied ideas — this will help deter­mine which is the supe­rior product.

That’s why we rec­om­mend a decen­tral­ized approach, where agen­cies need not con­form to a uni­form, cen­tral­ized out­look (or set of assump­tions). So we ask: why can’t the secure, encrypted national-​security intranet be more (a lit­tle more) like the inter­net and the blogosphere?

“Those who do not learn from his­tory are doomed to repeat it.”

—George San­tayana

Pref­ace: on Mon­day, we wrote Human Error (ver­sus Sys­temic Fail­ure), which sup­ple­mented our longer post from Sun­day: Intel­li­gence Fail­ures and Bad Infor­ma­tion Sys­tem Design. Much of that ‘Human Error’ post was devoted to men­tion­ing that within orga­ni­za­tions, most fail­ures, includ­ing human fail­ures, are sys­temic fail­ures. You can’t blame it on your sub­or­di­nates!

In the Sun­day post, we hypoth­e­sized and spec­u­lated that bad infor­ma­tion sys­tem design could be the cause of the recent intel­li­gence fail­ures. We based those sup­po­si­tions on our knowl­edge of infor­ma­tion sys­tems, com­mon design flaws, and the dys­func­tional nature of the fed­eral bureau­cracy but with no real or spe­cific knowl­edge of the cir­cum­stances. We don’t work for the gov­ern­ment, and we’re too lazy and too busy to inves­ti­gate on our own, but we fig­ured our hunches were cor­rect (and were will­ing to stake our mea­ger rep­u­ta­tion on them).

So, in the Mon­day post, we used L. Gor­don Crovitz’s col­umn, Intel­li­gence Is a Ter­ri­ble Thing to Waste, which appeared in that day’s edi­tion of The Wall Street Jour­nal, to pro­vide some anec­do­tal evi­dence to sup­port our con­jec­tures of the overly-​centralized and overly-​rigid nature of the systems.

We closed Monday’s post with: “Sad, but true.”

Unfor­tu­nately, an arti­cle in Friday’s edi­tion of The Wall Street Jour­nal, Years of Spotty Data-​Sharing on Sus­pects, pro­vides addi­tional evi­dence to sup­port much of the crit­i­cism that we levied on Sun­day (based upon our speculation).

We write “unfor­tu­nately,” because this is one of those cases where we hate to be right, but read it (the arti­cle) and weep. Here are sev­eral items men­tioned in the arti­cle and our comments.

Pres­i­dent Obama ordered agen­cies to bol­ster infor­ma­tion technology.

• It’s unlikely that the fail­ures are about tech­nol­ogy or inad­e­quate bud­gets. Note, using open-​source web apps, our database-​driven site and e-​mail costs less than $150 per year to oper­ate. It is a state-​of-​the-​art pub­lish­ing sys­tem that could be eas­ily used by depart­ments and agen­cies to post (and cat­e­go­rize) qual­i­ta­tive infor­ma­tion and leads. Those cat­e­gories could include sub­stan­ti­ated ver­sus unsub­stan­ti­ated claims.
• More likely it’s about sys­tem design. We’re not under-​estimating the vol­ume of data for some agen­cies, but we are ques­tion­ing the need to cen­tral­ize its stor­age and man­age­ment. More on this below.

A pre­vi­ous inte­gra­tion attempt, appro­pri­ately called the Infor­ma­tion Inte­gra­tion Pro­gram, failed.

• Is any­one sur­prised by that result?
• We sus­pect it is overly-​rigid and centralized.
• We also sus­pect that if such an inte­gra­tion attempt were to ever suc­ceed, it would be imme­di­ately obso­lete–most likely because some such agency upgraded one of its data­bases, and it is no longer integrable.

Sup­pos­edly, another inte­gra­tion attempt won’t be com­plete for two years.

• Remem­ber: the last attempt failed. So, why believe the two-​year deadline?
• It likely involves many indus­tri­ous and very hard-​working con­sul­tants spin­ning around on the lit­tle ham­ster wheels and sweat­ing pro­fusely, but with no real chance of suc­cess. It would be a Greek Tragedy if it weren’t an Amer­i­can one.
• There are needs for large sys­tems, but we sus­pect far fewer than presumed.
• The issue isn’t how to accu­mu­late all infor­ma­tion and data, it is how to access infor­ma­tion as effi­ciently as pos­si­ble. So, why should a mid­dle­man aggre­gate it when indi­vid­ual agen­cies could pub­lish it and searchers (with proper clear­ance) could imme­di­ately find it.

Empha­sis on con­nect­ing e-​mail systems

• Please see our post, Inex­pen­sive but Valu­able Web-​based MIS, espe­cially the sec­tion, ‘E-​mail as the Cen­tral Ner­vous System.’ No need to repeat the argu­ment here, but e-​mail is an inef­fi­cient man­age­ment infor­ma­tion sys­tem. Bet­ter and inex­pen­sive sub­sti­tutes exist.
• Com­mu­ni­ca­tion should be about be about pub­lish­ing facts, spec­u­la­tions, and opin­ions, and let­ting oth­ers search those posts or reports (and/​or receive feeds of future ones).
• E-​mail is archaic for these pur­poses. We ask, dear reader: do you know any one of our sev­eral e-​mail addresses? Unless or are a friend or acquain­tance, no, you don’t. Yet you can read our cur­rent and past spec­u­la­tions and be auto­mat­i­cally informed of future ones.
• Why shouldn’t intel­li­gence ana­lysts, within their own com­mu­ni­ties, have the same capac­i­ties that you, dear reader, have through­out the world­wide com­mu­nity that is the web? Pro­vided you live in a free, uncen­sored soci­ety, you have the capa­bil­ity at lit­tle or no cost. You can search for items of inter­est and read and eval­u­ate them based upon your knowl­edge and per­spec­tive. You can think we’re a fool or not, but you can make that assess­ment your­self for your par­tic­u­lar prob­lem or need. Why shouldn’t ana­lysts be able to do the same on their intranet?

National Intel­li­gence Library per­mits searches of fin­ished reports

• That’s good, but it’s not enough.
• How much sub­jec­tive and unsub­stan­ti­ated and unver­i­fied data are elim­i­nated from those fin­ished reports? Again, that’s the stuff of new leads and threat identifications.
• How long does it take for such reports to be “fin­ished” and avail­able for gen­eral consumption?
• If agen­cies or work groups had their own (secure, intranet) pub­lish­ing plat­forms, why bother con­sol­i­dat­ing? Let poten­tial users, with the right clear­ances, surf. Another way to ask: why bother con­sol­i­dat­ing when the con­sol­ida­tor can­not nec­es­sar­ily antic­i­pate the needs of users? Also, each blog on the web has its own sys­tem of per­mis­sions for access to pri­vate and password-​protected infor­ma­tion. Has any­one inves­ti­gated whether a cen­tral clear­ing­house is more effi­cient than main­tain­ing access to data at local lev­els. We don’t have many sub­scribers, but we know when we have new ones, and can grant var­i­ous lev­els of per­mis­sions to them.

Prob­lems search­ing unprocessed infor­ma­tion, espe­cially clearances

• See Sun­day or Monday’s post.
• Regard­ing who has access to which data­bases, secu­rity clear­ances are a major issue for a vari­ety of good rea­sons, but dis­tinc­tions should be made between data about cit­i­zens and for­eign­ers, and there is no rea­son to endow for­eign­ers with our rights; so, infor­ma­tion about for­eign­ers should be more eas­ily accessed.

Secu­rity clearances

• Obvi­ously nec­es­sary, clearly a con­straint. In fact, by def­i­n­i­tion, they are con­straints on sharing.
• We don’t have an answer to this issue, but we do have ques­tions: Is clear­ance a status-​symbol? Should lower level inves­ti­ga­tors and ana­lysts have greater access? What are the costs and ben­e­fits of greater access? How could leaks com­pro­mise var­i­ous inves­ti­ga­tions? Obvi­ously, records of vis­its, queries, etc, can be kept (just like we have at our site and most other web pub­lish­ers have).

Ter­ror­ist Iden­ti­ties Data­mart Environment

• We ask: who, other than a gov­ern­ment (or cor­po­rate) bureau­crat (or par­a­sitic con­sul­tant), could like that name? Seriously? Is it that cru­cial to cre­ate a word from the acronym?
• Does the man­gling of Eng­lish imply any­thing about the con­struc­tion of the sys­tem? We wonder.
• It’s clearly a cen­tral­ized sys­tem, and based upon the Crovitz col­umn we men­tioned above, it seems very dif­fi­cult to get on the list. We sus­pect it is harder to get off of the list.

What would we do?

For cer­tain stan­dard­ized mon­i­tor­ing and detec­tion sys­tems, there is a clear need for large data­bases. These are sim­i­lar to record-​keeping sys­tems for trans­ac­tions and events, i.e., not much dif­fer­ent than, say, keep­ing track of check­ing account trans­ac­tions or pur­chases and returns at Wal­Mart. In a world-​wide endeavor like ter­ror­ist detec­tion and mon­i­tor­ing, such sys­tems need to be search-​able web appli­ca­tions (on a pri­vate intranet). That very much reduces the need for con­sol­i­da­tion into one ginor­mous database.

In fact, the web is noth­ing if not one, large, search­able data­base (made of mil­lions of small ones). How­ever, the con­sol­i­da­tion and aggre­ga­tion is inher­ent and organic, rather than com­manded or centrally-​planned. In fact, mod­ern sites are database-​driven, and a visit to a page is the call to an (actual) data­base. Every time a Google search is per­formed, the web surfer is run­ning a query, and has access to some sites but not other, password-​protected ones.

More­over, the search engines have devel­oped algo­rithms to present the results in par­tic­u­lar ways, and they are incred­i­bly good at it. (At least on those searches where we rank high.) That is where time and effort should be devoted – not in attempt­ing to phys­i­cally con­sol­i­date dis­parate databases.

In that respect, let the dis­par­i­ties grow so that each agency can best serve its own mis­sion, yet pro­duce and pub­lish intranet-​accessible reports and notes.

We’d imag­ine that many of inves­ti­ga­tions are ad hoc and involve a bit of serendip­ity. We would imag­ine that with slightly dif­fer­ent mis­sions, the agen­cies have slightly dif­fer­ent data and infor­ma­tion require­ments and emphases and tra­di­tions and cul­tures. So, why try to cen­trally con­sol­i­date (and there­fore homog­e­nize) the unique sys­tems that may have evolved for spe­cific and good reasons.

How­ever, small, idio­syn­cratic sys­tems that com­prise a secu­rity intranet, can be index-​able and search-​able – just like the web.

So, we say har­ness the power of exist­ing web appli­ca­tions and tech­nol­ogy to pro­tect our nation. Allow inves­ti­ga­tors and ana­lysts be entre­pre­neur­ial pub­lish­ers of their idio­syn­cratic views, facts, and sup­po­si­tions. (All pri­vate and all secure on an intranet.)

Let inves­ti­ga­tors and ana­lysts pub­lish their reports and spec­u­la­tions for them­selves and other agen­cies, join forums, and con­verse with their col­leagues – even anony­mously. (We reit­er­ate: all pub­lished securely and pri­vately on a huge intranet, of course.) Let them use their intel­lects and train­ing to behave entre­pre­neuri­ally, not bureaucratically.)

Use cen­tral resources to develop search algo­rithms and secu­rity clearance/​permissions appli­ca­tions that oper­ate seam­lessly in a secure envi­ron­ment. Inte­grate intel­li­gently, not by con­sol­i­da­tion, by query. User man­age­ment and per­mis­sions are immensely impor­tant, but mil­lions of sites have solved such prob­lems. With a bit of guid­ance and in time, we think the gov­ern­ment can, too.

Infor­ma­tion: it’s like the econ­omy (and wealth) stu­pid. Try to cen­tral­ize it, and you’ll kill it and destroy the incen­tives to pro­duce more. In that respect, see The Wall Street Journal’s Review & Out­look, ‘A Fail­ure to Con­nect the Dots’, for more cor­rob­o­rat­ing evi­dence and perspective.

We’ll likely edit this post in the morn­ing. (We did, and will likely do so again.)

Is There a Dif­fer­ence? Sometimes.

After inter­mit­tently pon­der­ing the attempted Christ­mas Day bomb­ing of North­west Flight 253, yes­ter­day we pub­lished Intel­li­gence Fail­ures and Bad Infor­ma­tion Sys­tem Design.

Per its title, we spec­u­lated that bad infor­ma­tion sys­tem design could have been the cause of the fail­ure. In par­tic­u­lar, if the sys­tem is overly-​centralized and overly-​rigid intel­li­gence fail­ures can occur. (It’s a long post, but we think that it is well worth reading.)

Shortly after pub­lish­ing it last night, we saw today’s (Jan 4) opin­ion col­umn by The Wall Street Journal’s L. Gor­don Crovitz. His col­umn is enti­tled Intel­li­gence Is a Ter­ri­ble Thing to Waste.

In it, he quotes the head of the FBI’s Ter­ror­ist Screen­ing Cen­ter, Tim­o­thy Healy, and Mr. Healy’s expla­na­tion of “rea­son­able sus­pi­cion,” which is what it takes to get on a “list.”

“Rea­son­able sus­pi­cion requires ‘artic­u­la­ble’ facts which, taken together with ratio­nal infer­ences, rea­son­ably war­rant a deter­mi­na­tion that an indi­vid­ual is known or sus­pected to be or has been engaged in con­duct con­sti­tut­ing, in prepa­ra­tion for, in aid of, or related to, ter­ror­ism and ter­ror­ist activ­i­ties, and is based on the total­ity of the cir­cum­stances. Mere guesses or inar­tic­u­late ‘hunches’ are not enough to con­sti­tute rea­son­able suspicion.”

Mr. Crovitz then goes on to explain that if Mr. Healy’s expla­na­tion sounds like legalese that’s because it is and that it is silly and dan­ger­ous (our words) to treat poten­tial for­eign ter­ror­ists and enemy com­bat­ants as domes­tic criminals.

That’s very sim­i­lar to we wrote yesterday:

…In par­tic­u­lar, we could imag­ine that unver­i­fied and unsub­stan­ti­ated reports are among the least generally-​accessible data – until they are ver­i­fied, reviewed or accepted by the bureau­cracy, regard­less of whether that involves a sin­gle agency or an over-​seeing umbrella group.

BUT those unsub­stan­ti­ated reports are the ones that are most likely to pro­vide infor­ma­tion about new ter­ror­ists like Abdul­mu­tal­lab, (and that is the prob­lem with treat­ing for­eign­ers who are threats to our national secu­rity as crim­i­nals rather than enemy com­bat­ants.) If our hunch is cor­rect, then one should expect future “intel­li­gence fail­ures” to arise in sim­i­lar situations.

More­over, if our hunch is cor­rect, then a cen­tral­ized, data­base administrator’s (rather arbi­trary) rules – or worse, some lawyer’s rules – sub­sti­tute for the indi­vid­ual knowl­edge and dis­cre­tion of var­i­ous field agents and super­vi­sors.³ As such, fields agents may not have the oppor­tu­nity to syn­the­size the infor­ma­tion until it is too late. (It’s a case of the per­fect being the enemy of the good.)

By the way, Mr. Crovitz con­cludes with:

We have a choice. We can limit how infor­ma­tion is used or we can allow smart use of infor­ma­tion to pre­vent attacks. If we con­tinue to choose to limit how infor­ma­tion can be used in our defense, we shouldn’t be sur­prised when our defenses fail.

In that clos­ing para­graph, he suc­cinctly states the prob­lem that we more pre­cisely explain (in terms of infor­ma­tion sys­tem design) and our rec­om­men­da­tion to make security-​related infor­ma­tion sys­tems more like the inter­net and blogosphere.

When Human Errors Are Sys­temic Errors

Please note our foot­note (#3) in the third para­graph of the above excerpt from yesterday’s post. It reads: In this post, we won’t pro­vide any sup­port for the fol­low­ing state­ment , but, like errors in bank­ing and the finan­cial ser­vices (and almost every­thing else), we pre­fer errors to be idio­syn­cratic rather than sys­temic. (See Sys­temic Risk Reg­u­la­tion and Irony and espe­cially Idio­syn­cratic and Con­cen­tra­tion Risk, Again for our per­spec­tive in those areas.)

We noticed an arti­cle in today’s Pitts­burgh Post-​Gazette, Bomb attempt blamed on human error that describes Deputy national secu­rity adviser John Brennan’s expla­na­tion for the secu­rity fail­ure.¹ To para­phrase, he said it was human error.

Blam­ing some­thing on “human error” makes it seem like an indi­vid­ual, rather than the sys­tem, failed – like the sole check­point oper­a­tor arrived late because he was hung-​over from a Christ­mas party. However, unless some device or dog fails, all errors are human errors. In fact, one could argue that device and canine errors are human errors, too, because the plan­ner or designer did not have the fore­sight to antic­i­pate and mit­i­gate those errors or failures.

So, one – that would be us – could argue that fix­ing the blame on human error isn’t very descrip­tive or use­ful. If at some level, all such errors are human errors, then we haven’t been told much or learned much. We don’t know if those “human errors” are truly idio­syn­cratic or sys­temic. Did a poorly-​designed sys­tem induce higher lev­els (or prob­a­bil­i­ties) of human errors (than what could have been)? We don’t know.

When it seems rea­son­able to assume that near-​perfect detec­tion is demanded, we won­der why the sys­tem designer or admin­is­tra­tor would per­mit truly idio­syn­cratic errors, and we won­der if con­tin­gen­cies have been devel­oped in case of failures.

We’re not call­ing for over-​complicated solu­tions, just a lit­tle fore­sight. In that sense, it’s not dif­fer­ent than plan­ning for the failure-​related activ­i­ties in man­u­fac­tur­ing or any other field of endeavor. Such plan­ning should occur ex ante, rather than ex post, but does it?²

So, unless there was egre­gious, crim­i­nal, or trea­so­nous behav­ior by a mem­ber of one of our secu­rity forces, blam­ing human error doesn’t answer the ques­tion of what went wrong, and does little-​to-​nothing to pre­vent such prob­lems in the future. More­over, it val­i­dates what we wrote yes­ter­day (imme­di­ately below the excerpts shown above):

Unfor­tu­nately, that prob­lem is exac­er­bated once those rules and poli­cies are set. Later admin­is­tra­tors may be unwill­ing to “rock the boat” and ini­ti­ate worth­while changes because there is a chance of being blamed for sub­se­quent fail­ures but lit­tle chance of being rewarded for success. (Those acco­lades would most likely go to the “eagle-​eyed” agent who noticed some­thing was wrong.) By the way, as we often argue, it is dif­fi­cult to cat­e­go­rize such a choice – not to act – as irre­spon­si­ble behav­ior, espe­cially when it is induced by poorly-​designed poli­cies and a lack of man­age­r­ial dis­ci­pline. That’s why it is a bureau­cracy, after all.

So, rigid poli­cies self-​perpetuate and infor­ma­tion, hunches, and rumors are not passed along.

Sad, but true.

Update: What tim­ing! Moments after we pub­lished this, we saw this col­umn, Intel­li­gence Is a Ter­ri­ble Thing to Waste, by L. Gor­don Crovitz at The Wall Street Journal’s web site. It nicely com­ple­ments our post and val­i­dates a few of our spec­u­la­tions – although we must admit that his col­umn has a catch­ier title.

In this rather long post we spec­u­late about a pos­si­ble under­ly­ing cause of the “intel­li­gence fail­ure” involv­ing Umar Farouk Abdul­mu­tal­lab, the Niger­ian accused of try­ing to blow-​up North­west Flight 253 on Christ­mas Day. Of inter­est is how he was cleared to fly despite his father noti­fy­ing U.S. author­i­ties of his – the son’s – extrem­ism and poten­tial for terrorism.

Note that we have absolutely no pri­vate infor­ma­tion regard­ing either the inci­dent or gov­ern­ment infor­ma­tion sys­tems; so, we spec­u­late based upon our knowl­edge of other large, bureau­cratic orga­ni­za­tions with rigid, poorly-​designed systems.

We real­ize that incen­tive prob­lems – which result in the unwill­ing­ness of agen­cies and indi­vid­u­als to share data and infor­ma­tion across juris­dic­tions – and our free­doms and rights con­strain the effec­tive­ness of inves­tiga­tive efforts, but for the most part, we’ll ignore those issues to focus on infor­ma­tion systems.

Com­mon MIS Issues & Problems

A few weeks ago we wrote Inex­pen­sive but Valu­able Web-​based MIS. Besides describ­ing those ben­e­fi­cial sys­tems, we men­tioned that many so-​called “man­age­ment infor­ma­tion sys­tems” are, in fact, merely data-​processing and record-​keeping sys­tems (for trans­ac­tions and events).

Such sys­tems rarely pro­vide infor­ma­tion – decision-​altering con­tent – for the types of strate­gic deci­sions made by senior man­agers, and unfor­tu­nately, they may not be well-​designed to pro­vide use­ful tac­ti­cal infor­ma­tion, either. That’s the case if the systems:

1. Pro­duce use­less stan­dard­ized out­put (reports);
2. Are dif­fi­cult to fully access or query; or
3. Don’t adapt quickly or well to changes in the envi­ron­ment, oper­a­tions or insti­tu­tional knowledge.

In Umar Farouk Abdulmutallab’s case, we sus­pect that it is the inher­ent rigid­ity of the data­base appli­ca­tion and/​or the rigid­ity of the design­ers’ thought processes that are to blame. (Note that for new infor­ma­tion sys­tems, use­less stan­dard­ized reports result when sys­tems design­ers don’t ask users the cor­rect ques­tions or do ask the right ques­tions, but don’t really under­stand the replies. See Details Are Not Infor­ma­tion for more on this topic. One of our MIS friends often remarks that her key func­tion is to serve as a trans­la­tor between sys­tem users and sys­tem devel­op­ers, and that role is crit­i­cal but too often ignored. For older sys­tems, irrel­e­vance and obso­les­cence usu­ally result when the sys­tem isn’t easy to change.)

What Went Wrong on Christmas?

When bad things hap­pen, i.e., when some­one like Umar Farouk Abdul­mu­tal­lab squeezes through the detec­tion sieve, it is pos­si­ble that noth­ing failed. One must con­sider that the detec­tion sys­tem – the net, the fil­ter, the web – may not have designed to catch every­thing and that the designer or owner con­sid­ered a cer­tain level of error or mis­clas­si­fi­ca­tion to be accept­able. The designer may have con­cluded that a per­fect, error-​free sys­tem is too expen­sive to develop and main­tain.¹

How­ever, the fail­ure in the Abdul­mu­tal­lab case was so egre­gious that it seems far more likely that either the detec­tion sys­tem was either incom­pe­tently designed or administered.

Now, it is quite pos­si­ble that a gov­ern­ment sen­try or sen­tinel fell asleep or neglected his or her respon­si­bil­ity. In that case, it is both a human error – because a per­son failed – but also a sys­temic error because there was no redun­dancy or backup mit­i­gate such error. How­ever, rather than crit­i­cize gov­ern­ment employ­ees involved with the nation’s secu­rity, we’ll assume that they are earnest, capa­ble, and hard-​working as we believe that is true.

In that case, it must be that despite their best efforts, the detec­tion sys­tem failed, and one rea­son for the fail­ure could be the improper design of the government’s infor­ma­tion system.

One obvi­ous weak­ness in the ter­ror­ist detec­tion sys­tem – and it is by design – is the government’s unwill­ing­ness to use con­di­tional prob­a­bil­i­ties to assess the like­li­hood that some­one is a ter­ror­ist, espe­cially if the per­son is a for­eigner and is not pro­tected by our Con­sti­tu­tion and Bill of Rights. As we wrote in The Absur­dity of Has­sling Grandma but not Nidal Hasan, we do blame the gov­ern­ment (and Pres­i­dent Obama) for main­tain­ing poli­cies and pro­ce­dures that ignore infor­ma­tion, i.e., prior and pos­te­rior (con­di­tional) prob­a­bil­i­ties that some­one fits the well-​defined pro­file of a terrorist.

How­ever, other than crit­i­ciz­ing his unwill­ing­ness to “pro­file,” we don’t blame Pres­i­dent Obama for the fail­ure on Christ­mas, and we think that it is silly for oth­ers to blame him.

We do think that his pref­er­ences and mind­set for large, cen­tral­ized, mech­a­nisms – e.g., nation­al­ized health-​care, bail-​outs, etc – are sim­i­lar to the prob­lem we dis­cuss below, but in all like­li­hood, the sys­tem pre­dates his tenure.²

So, despite the sys­tem hand­i­capped by the unwill­ing­ness to pro­file, if the intel­li­gence fail­ure was not Pres­i­dent Obama’s fault (and not for­mer Pres­i­dent Bush’s fault) and it is not the fault of those man­ning the sys­tems, than who or what is to blame? We sug­gest that the reader con­sider a poorly-​designed, overly-​rigid database/​information system.

Too Rigid

By def­i­n­i­tion, in an overly-​rigid infor­ma­tion sys­tem, both the input and out­put func­tions may be less flex­i­ble and user-​friendly than required. Given the fed­eral government’s pen­chant for large, cen­tral­ized, stan­dard­ized solu­tions, it is easy for us to believe that such an infor­ma­tion sys­tem (or sys­tems) has (have) been employed in the war against ter­ror­ism and that such sys­tems increase the like­li­hood of “intel­li­gence fail­ures” and ter­ror­ists evad­ing detection.

Rigid Input: Round Holes, Square Pegs and Worse

Con­sider the idiom of “putting a square peg in a round hole.” For data­bases that means that cer­tain facts that should be recorded may not be eas­ily cat­e­go­rized into avail­able fields because proper, descrip­tive fields do not exist (and can­not be eas­ily added). For exam­ple, con­sider cen­sus or EEOC forms where there is no appro­pri­ate box to check: where it is required to select a sin­gle “nation­al­ity” or “race” when you are ¹⁄₁₆ of this and ¹⁄₈ of that, et. al.

If such metaphor­i­cal “square pegs” could con­sis­tently be jammed into “round holes,” there would not be an issue because users would likely have devel­oped heuris­tics (rules-​of-​thumb) to cre­ate well-​formed sub­sti­tu­tions and work-​arounds. In all like­li­hood, those rules or map­pings would not be for­mal­ized in any offi­cial man­ual or doc­u­men­ta­tion, but they would be well-​known and trans­mit­ted dur­ing both for­mal and infor­mal train­ing sessions.

Unfor­tu­nately, real-​life is often not so sim­ple, because the so-​called “square pegs” may not be of, say, uni­form size, color, and shape.

In fact, other than cer­tain fields like names and addresses, we sus­pect that many of the facts that should be recorded can’t be eas­ily or suc­cinctly described in a word or two – that they are more nuanced and qual­i­ta­tive and grad­u­ated and require length­ier, usu­ally sub­jec­tive descrip­tions. Actu­ally, they may not be very dif­fer­ent than blog posts, and we would hope that writ­ers and recorders of those posts would have the flex­i­bil­ity to cre­ate new fields and cat­e­gories on-​the-​fly – like we do every time we add a new tag or category.

Unfor­tu­nately, we sus­pect that leads to many “cod­ing” errors and incon­sis­ten­cies and extremely long descrip­tions of fields (to pre­vent such “errors”.) We also sus­pect that it leads to too much over­sight; many lay­ers of approval by supe­ri­ors (and there­fore much edit­ing and chang­ing); and overly-​restrictive input poli­cies, e.g., “he doesn’t have the per­mis­sion or author­ity to write that.”

More­over, we also sus­pect that these prob­lems are exac­er­bated when inves­ti­ga­tors and field agents aren’t involved in the infor­ma­tion sys­tem design process.

Rigid Out­put

Other prob­lems with rigid, poorly-​designed sys­tems include (1) not pro­vid­ing use­ful, stan­dard­ized out­put or (2) not hav­ing the capac­ity for users to eas­ily search and access stored data for ad hoc queries.

Note again that we have no knowl­edge of actual, rou­tine TSA, FBI, CIA, and Home­land Secu­rity reports, and if we did, we prob­a­bly couldn’t write anything.

1. Too Cen­tral­ized and Uniform

That being said, we could imag­ine that there are dif­fer­ent lev­els of secu­rity clear­ance, and that access to the data could be overly-​restricted based upon those clear­ances. In par­tic­u­lar, we could imag­ine that unver­i­fied and unsub­stan­ti­ated reports are among the least generally-​accessible data – until they are ver­i­fied, reviewed or accepted by the bureau­cracy, regard­less of whether that involves a sin­gle agency or an over-​seeing umbrella group.

BUT those unsub­stan­ti­ated reports are the ones that are most likely to pro­vide infor­ma­tion about new ter­ror­ists like Abdul­mu­tal­lab, (and that is the prob­lem with treat­ing for­eign­ers who are threats to our national secu­rity as crim­i­nals rather than enemy com­bat­ants.) If our hunch is cor­rect, then one should expect future “intel­li­gence fail­ures” to arise in sim­i­lar situations.

More­over, if our hunch is cor­rect, then a cen­tral­ized, data­base administrator’s (rather arbi­trary) rules – or worse, some lawyer’s rules – sub­sti­tute for the indi­vid­ual knowl­edge and dis­cre­tion of var­i­ous field agents and super­vi­sors.³ As such, fields agents may not have the oppor­tu­nity to syn­the­size the infor­ma­tion until it is too late. (It’s a case of the per­fect being the enemy of the good.)

Unfor­tu­nately, that prob­lem is exac­er­bated once those rules and poli­cies are set. Later admin­is­tra­tors may be unwill­ing to “rock the boat” and ini­ti­ate worth­while changes because there is a chance of being blamed for sub­se­quent fail­ures but lit­tle chance of being rewarded for success. (Those acco­lades would most likely go to the “eagle-​eyed” agent who noticed some­thing was wrong.) By the way, as we often argue, it is dif­fi­cult to cat­e­go­rize such a choice – not to act – as irre­spon­si­ble behav­ior, espe­cially when it is induced by poorly-​designed poli­cies and a lack of man­age­r­ial dis­ci­pline. That’s why it is a bureau­cracy, after all.

So, rigid poli­cies self-​perpetuate and infor­ma­tion, hunches, and rumors are not passed along.

2. Search­able? We Doubt It.

As we have repeat­edly men­tioned, much of this post is mere spec­u­la­tion. A few of our con­jec­tures are pro­jec­tions based upon our own expe­ri­ences. Given that, we could imag­ine that inves­ti­ga­tors, ana­lysts, and agents can­not query or search the entire data­base (if it exists in one place).

Most likely, they receive exported sub­sets of the data, and those sub­set do not arrive imme­di­ately upon request. (The deci­sion to grant the request is prob­a­bly made by a data­base man­ager or admin­is­tra­tor and may require detailed spec­i­fi­ca­tions and pos­si­bly mul­ti­ple approvals – a whole process. Again, that’s why it is a bureaucracy.)

Now, we’re not sure of the ben­e­fits of such a bureau­cracy and sus­pect that such processes con­tinue to exist because “that’s how we’ve always done it,” which could be trans­lated as “we don’t know any better.”

Regard­less, there are costs to such pro­ce­dures. Besides the pos­si­ble lack of time­li­ness, there is a reduced oppor­tu­nity of dis­cov­er­ing any­thing – pat­terns, what not – acci­den­tally or serendip­i­tously. When a sub­set or export is requested and jus­ti­fied it must be com­pletely spec­i­fied; so, the requester needs to know exactly what he or she plans to inves­ti­gate before com­plet­ing a request and there is lit­tle chance of expand­ing or redi­rect­ing the inves­ti­ga­tion with­out re-​submitting requests for addi­tional fields.

In addi­tion, if the entire data­base is not fully-​searchable, then inves­ti­ga­tors are less likely to find matches and pat­terns across fields. Recall our crit­i­cism above: with rigid input fields, and vary­ing “square pegs,” agents in dif­fer­ent loca­tions and depart­ments may input sim­i­lar facts in dif­fer­ent fields. If some of those fields are not avail­able and search­able, then inves­ti­ga­tors will get fewer hits and matches and that will reduce the chance of mak­ing con­nec­tions and discoveries.

Our Rec­om­men­da­tion

So. the dili­gent reader, who has made it to this point, may ask: if your hypothe­ses and spec­u­la­tions are cor­rect, then what’s your solu­tion? (Alter­na­tively, they may note that the sell­ers of ham­mers tend to see a lot of nails.)

We reply with a rhetor­i­cal ques­tion: why can’t such sys­tems or con­glom­er­a­tions of sys­tems be more like the web and blo­gos­phere? By that we mean why can’t they be unfet­tered, completely-​searchable, accept respon­si­ble com­ments and ques­tions, and even per­mit writ­ers with vary­ing degrees of cred­i­bil­ity to post entries. (If the gov­ern­ment already has such a sys­tem, then kudos to it.)

Why not decen­tral­ize the process and empower secu­rity inves­ti­ga­tors, ana­lysts, and agents to use their idio­syn­cratic beliefs, opin­ions, infor­ma­tion, expe­ri­ences, posi­tions, and knowl­edge to iden­tify prob­lems and to adapt the data­base as threats and knowl­edge change?⁴

We imag­ine a mini-​version of the inter­net (with the abil­ity to search the entire inter­net, too), where indi­vid­ual agen­cies pub­lish blogs and news reports for them­selves and other agen­cies. (Geez, they could even sign-​up for each oth­ers’ feeds.

Of course, such a sys­tem would need to be at least as secure as on-​line bank­ing, but more pri­vate, but all such sys­tems must be.

Note, also, that noth­ing pre­cludes the run­ning or har­vest­ing of rou­tine reports from such sites. That’s what search engines and their bots and a host of sites already do. They stan­dard­ize the out­put of many dis­parate sys­tems. In fact, our rec­om­men­da­tion does not require any new or advanced tech­nol­ogy – just the appli­ca­tion of exist­ing plat­forms that are freely and read­ily avail­able to any­one with a few bucks and an inter­net connection.

Granted, it’s on a much larger scale than our blog, but it need not be expen­sive.⁵ More­over, we sus­pect that access to exist­ing sys­tems could be incor­po­rated eas­ier via web apps than through cus­tom pro­gram­ming for­ays that attempt to merge or con­sol­i­date exist­ing data­bases. For exam­ple, every Google query searches mil­lions of MySQL and MSSQL data­bases all with slightly dif­fer­ent struc­tures and fields.

Maybe we’re wrong, maybe we’re right. How­ever, even if our diag­no­sis is cor­rect, we doubt that the gov­ern­ment would act on our rec­om­men­da­tion. It would most likely try a cen­tral­ized “fix” of the iden­ti­fied prob­lems or would try a pilot-​program that (due to its lim­ited nature) would be des­tined to fail. In that case, hop­ing for con­tin­ued good luck might be the most rea­son­able and viable strategy.

In clos­ing, note that we are not dis­parag­ing the efforts of our fel­low cit­i­zens or the nation’s allies in their defense of our coun­try and way of life. Instead, if our spec­u­la­tions are cor­rect (or nearly so) we are rec­om­mend­ing a change in strat­egy and tac­tics so that their earnest effort yields more pro­duc­tive results.

As usual with long posts, we’ll likely make cor­rec­tions and edits that clar­ify our prose dur­ing the next few days.

Copy­right © 2010 Spero Consulting.

---

Foot­notes:

The Joy of the Season… is Lost on Him

The chil­dren are all snug in their beds with visions of Wii’s danc­ing in their heads, and after four weeks, we finally have the chance to fin­ish a post that we started on Thanks­giv­ing Day, when we read a very silly col­umn by George Will.

His col­umn was enti­tled, “No gifts, please.” In it he dis­cusses (and approves of) a pam­phlet called Scrooge­nomics: Why You Shouldn’t Buy Presents for the Hol­i­days by Joel Wald­fo­gel. Based upon both Mr. Will’s col­umn and the book’s title, it seems that Mr. Wald­fo­gel believes that buy­ing Christ­mas gifts destroys value, i.e., the col­lec­tive sat­is­fac­tion in the econ­omy (and per­haps the world) would be greater if Christ­mas gift-​givers spent the money on them­selves instead of on loved ones and others.

Before crit­i­ciz­ing that con­clu­sion, we do want to men­tion a few caveats:

1. We’re not sure if Mr. Will wrote the col­umn in a tongue-​in-​cheek man­ner, but there is no indi­ca­tion that he is any­thing but seri­ous, and the col­umn was pub­lished on Thanks­giv­ing, not April Fool’s Day.¹ Also, we do real­ize that he is con­strained to only a cer­tain num­ber of words per col­umn, and that rel­a­tively low num­ber may make com­plete and clear com­mu­ni­ca­tion dif­fi­cult. Or, maybe a dim edi­tor unknow­ingly hacked the wit out of the column.
2. We didn’t and don’t have the time or inter­est to locate, buy, and read Mr. Waldfogel’s book, so it may be more nuanced and sophis­ti­cated than Mr. Will under­stands it or cares to explain it. If that’s the case, then we apol­o­gize to Mr. Wald­fo­gel for group­ing him with Mr. Will.
3. Sim­i­larly, Mr. Wald­fo­gel may have writ­ten the book in either a tongue-​cheek-​manner or – using rhetoric and exag­ger­a­tion – in an instruc­tive way to edu­cate his stu­dents. If that’s the case then, again, we are sorry for crit­i­ciz­ing him, and we note that Mr. Will did him no favor by pub­li­ciz­ing his work in such a poor manner.

That being said, using com­mon sense and a lit­tle knowl­edge of eco­nom­ics, there are at least four ways to negate their conclusions:

1. Revealed pref­er­ence of gift-​giving actions..
2. The pref­er­ences, includ­ing the risk pref­er­ences, of gift givers and receivers and what that means for col­lec­tive satisfaction.
3. The exis­tence of uncer­tainty as it relates to the level of enjoyment.
4. Whether one truly knows ones own tastes and how they will change through time and if oth­ers know better.

The first three can be dis­cussed within the con­text of stan­dard micro­eco­nom­ics and util­ity the­ory, while the fourth one involves the vio­la­tion of one of the most com­mon and basic assump­tions in economics.

Revealed Pref­er­ence

This is the sim­plest argu­ment to fol­low. Revealed pref­er­ences means that actions speak louder than words – i.e., what peo­ple do is what mat­ters – not what they say.

The fact that ratio­nal folks vol­un­tar­ily buy presents for oth­ers means that they have con­cluded it is the best use of their money (or credit). Oth­er­wise, one must assume that those being stud­ied are irra­tional and/​or that fac­tors other than their own con­sump­tion affect their well-​being or sat­is­fac­tion. Both are easy to believe, but once an econ­o­mist assumes irra­tional­ity it becomes very dif­fi­cult to draw any conclusions.

In fact, if they are irra­tional, telling them so isn’t likely to have much of an effect on them or their behavior.

For exam­ple, if folks are irra­tional, then what can Mr. Wald­fo­gel say? “Many peo­ple can’t think clearly and waste money.” We have no doubt that it is true, but that’s not very insight­ful or new. In that case, they don’t just waste money at Christ­mas but dur­ing every other sea­son. (Per­haps they waste less at Christ­mas than they oth­er­wise would (with­out the empha­sis on peace, love, and joy and shop­ping for others).

Now it is pos­si­ble that Mr. Wald­fo­gel has a game the­o­retic model – sim­i­lar to, say, the Prisoner’s Dilemma – in mind, where some “bad” equi­lib­rium of gift-​giving obtains by every­one behav­ing ratio­nally. But, for a vari­ety of rea­sons, every Christ­mas peo­ple stop buy­ing presents for other folks. So, unlike a text­book prob­lem that involves incar­cer­a­tion or com­mit­ment, it seems that real peo­ple can and walk away in the mid­dle of the “game” if they want.

Thus, when he states that value is destroyed, given the fact that folks con­tinue to do it, it may sim­ply be the fact that Mr. Wald­fo­gel may have mis-​measured the over­all sat­is­fac­tion folks get from giv­ing and receiving.

Risk Pref­er­ences

Pre­sum­ably, to con­clude that value is destroyed by Christ­mas gift giv­ing – and to assign a dol­lar amount to it – some mea­sure of sat­is­fac­tion, like util­ity, must have been assumed – at least implic­itly. More­over, that mea­sure, e.g., util­ity func­tion, must be invert­ible (from sat­is­fac­tion to dollars) .

Again, with­out read­ing the book, there seem to be a num­ber of ways to defeat the asser­tion that value is destroyed. Here are a few.

First, col­lec­tive sat­is­fac­tion can be aggre­gated and weighed any arbi­trary num­ber of ways. For every set­ting that shows value is destroyed, one can weigh the sat­is­fac­tion of recip­i­ents in such a way that val­ued is created.

Sec­ondly, obvi­ously indi­vid­u­als may have dif­fer­ent pref­er­ences for dif­fer­ent goods and ser­vices, and they may have dif­fer­ent pref­er­ences for the same good or ser­vice at dif­fer­ent lev­els of (their own) wealth and con­sump­tion. Most eco­nomic mod­els keep things sim­ple and assume that indi­vid­u­als have rel­a­tively con­stant pref­er­ences, i.e., that they are either risk-​neutral, risk-​loving, or risk-​averse for the entire range of consumption.

For a sim­ple world where there is only one good, econ­o­mists gen­er­ally assume that sat­is­fac­tion increases (or at least doesn’t decrease) as more of that good is con­sumed. So, in a “sin­gle period,”

• Risk-​neutral means that pref­er­ences are lin­ear with respect to the good. That means that regard­less of the num­ber of units con­sumed, each addi­tional unit brings the same level of sat­is­fac­tion, i.e., the 31^st ham­burger is as sat­is­fy­ing as the first.
• Risk-​loving means that pref­er­ences are con­vex with respect to the good, i.e, the more units con­sumed, the greater the mar­ginal sat­is­fac­tion. It is very rare to see such behav­ior. It means that in a world con­sist­ing only of ham­burg­ers, eat­ing the 31^st burger brings more sat­is­fac­tion than eat­ing any pre­vi­ous one, includ­ing the first one.
• Risk-​averse means that pref­er­ences are con­cave with respect to the good, i.e., the more units con­sumed, the lesser the mar­ginal sat­is­fac­tion, i.e., the 31^st ham­burger eaten is the least sat­is­fy­ing – although still pos­i­tively satisfying.

Now it is pos­si­ble in multi-​period, over-​lapping gen­er­a­tion mod­els, for folks to get sat­is­fac­tion from both con­sum­ing and giv­ing – by fol­low­ing the Golden Rule of giv­ing this period in exchange for receiv­ing in the future, but that’s not what Christ­mas is about. The young don’t give presents to the old in hopes of get­ting presents when they are old.

With­out pre­sent­ing any­thing close to a for­mal model, con­sider a case where every­one is risk-​averse. If they have the same util­ity func­tion, then wealthy indi­vid­u­als for­sake less sat­is­fac­tion (from, say, the 31^st ham­burger) by giv­ing or donat­ing to less wealthy folks who gain by receiv­ing such ben­e­fits. (Remem­ber with similarly-​preferenced, risk – averse agents, wealthy folks give up less (mar­ginal) sat­is­fac­tion or util­ity than poorer gain. Draw an increas­ing, con­cave func­tion, like a square-​root func­tion to con­vince your­self that sat­is­fac­tion increases with con­sump­tion but at a decreas­ing rate.))

The last time that we checked, there were not a lot of mod­els that incor­po­rated dona­tions and giv­ing, but even if the net spenders lose some sat­is­fac­tion by con­sum­ing less due to their gifts – let’s say they are com­pelled to con­sume less or donate – then those acts do not destroy value if the recip­i­ents are par­tic­u­larly grate­ful and derive more sat­is­fac­tion from the gift than to givers lose by not behav­ing selfishly.

Of course, Mr. Will and Mr. Wald­fo­gel may be sur­prised that indi­vid­u­als enjoy giv­ing of them­selves and their resources – whether to fam­ily, friends, or strangers. The fact that such acts are out­side of many eco­nomic mod­els, doesn’t make such peo­ple irra­tional, it makes them kind and generous.

More­over, if econ­o­mists have a dif­fi­cult time mod­el­ing such behav­ior, that’s not a short-​coming of gen­er­ous indi­vid­u­als, it is a short-​coming of eco­nom­ics (and note that we are extremely sym­pa­thetic to the dif­fi­culty of mod­el­ing such social phenomena).

Now, par­si­mo­nious read­ers may dis­agree with our con­clu­sion that net sat­is­fac­tion is higher because of Christ­mas gift-​giving and exchang­ing because they may argue that the giver may not know what the recip­i­ent wants. We’ll rebut that posi­tion with two related, but sep­a­rate points: (1) in uncer­tain envi­ron­ments, you don’t always get what you want or pay for, and (2) in real-​life, you don’t always know what you want – i.e., what makes you happy or what’s best for you.

The Pres­ence of Uncertainty

Who knows? It may sim­ply be the case that no one cares enough about Mr. Will to have given him a present that he wanted, and that may have caused his Scrooge-​like col­umn, or maybe he was told one too many times that “you’ll shoot your eye out.”

That scene from A Christ­mas Story of the ric­o­chet­ing BB break­ing the Ralphie’s glasses per­fectly illus­trates that in real life, ran­dom, uncon­trol­lable events affect the enjoy­ment of any every good and ser­vice consumed.

So, when com­bined with bad luck, get­ting exactly what you want can be dis­as­trous, and far worse than get­ting a safer, less-​desirable ver­sion of the good or ser­vice or a sub­sti­tute. For exam­ple, think of the teenage boy who would pre­fer a sports car to a large sedan, but is far more likely to harm him­self and/​or oth­ers in a new Corvette as com­pared to a new Suburban.

So, before one can con­clude the net value is destroyed, one must be able to aggre­gate the actual sat­is­fac­tion received rather than the expected sat­is­fac­tion when the gift is received. In other words, it is impor­tant to know the range of out­comes and their prob­a­bil­i­ties – when they are know­able – to mea­sure how things worked out.

More pre­cisely, for risk-​averse indi­vid­u­als, it’s not just the mean that mat­ters, but the dis­tri­b­u­tion mat­ters, too, and oth­ers may under­stand (or have bet­ter infor­ma­tion about) that dis­tri­b­u­tion than the recip­i­ent. That means that their gifts may be more appro­pri­ate, and have a higher con­di­tional expected value.

Does that mean that oth­ers always give bet­ter presents than items pur­chased by ones self? Of course not, but it means that in some cases, the dif­fer­ence might be smaller than sup­posed or negative.

Know Thy­self, or Not

We’ll read­ily admit that our three counter-​arguments are not very pre­cise. How­ever, each implic­itly assumes that (at least) the recip­i­ents, know what’s best for them­selves.²

As we men­tioned above, with uncer­tainty (and dif­fer­ent pri­ors and sig­nals, which gen­er­ate dif­fer­ent pos­te­ri­ors) it is pos­si­ble that givers may have bet­ter knowl­edge of the envi­ron­ment, i.e., the dis­tri­b­u­tion of (pos­si­ble) out­comes, and, there­fore, they may be able to select the most appro­pri­ate gift for the recipient.

In this sec­tion, how­ever, we have some­thing a bit dif­fer­ent in mind. It is pos­si­ble that on cer­tain dimen­sions, oth­ers know you bet­ter than you know your­self. Thus, they may be able to select more appro­pri­ate (use­ful, val­ued, appre­ci­ated) gifts than the recip­i­ent could have selected for him or herself.

We sus­pect that like us, most read­ers didn’t always get what they asked for, but on some of those occa­sions ended up lik­ing the sub­sti­tute bet­ter than the orig­i­nal. We also sus­pect that sen­si­ble read­ers will agree that until a cer­tain age, par­ents def­i­nitely know bet­ter than their chil­dren what will max­i­mize the child’s expected, dis­counted, satisfaction.

The ques­tion is: at what age does that change? Is it when a twelve-​year-​old wants a Wii, or is into adult­hood, when the child wishes to marry. (In some cul­tures, with arranged mar­riages, it is at least that age.)

In most math­e­mat­i­cal mod­els of pref­er­ences, indi­vid­u­als know their own pref­er­ences. Oth­er­wise, it is dif­fi­cult to sole the indi­vid­u­als max­i­miza­tion prob­lem, i.e., what does one max­i­mize or optimized?

In real-​life that isn’t always true. More­over, pref­er­ences do change. Do you, dear reader, eat the exact same foods that you did as an infant; as a high-​metabolism teenager; as a col­lege stu­dent? While some folks are sur­pris­ingly con­sis­tent, most aren’t, and a favorite restau­rant one day is no longer vis­ited the next.

So, if one is will­ing to admit that one doesn’t always know what one will like in the future and one can­not com­pletely spec­ify how to order likes and dis­likes (pref­er­ences) today, is it that much of a stretch to imag­ine that oth­ers may know cer­tain aspects of your pref­er­ences (and behav­ior) bet­ter than you, your­self, espe­cially how those pref­er­ences may evolve through time? We don’t think so.

Thus, is it that dif­fi­cult to believe that the net result of informed gift-​giving increases col­lec­tive hap­pi­ness despite the fact that many inap­pro­pri­ate gifts are received (and regifted)?

Merry Christ­mas!

Now we men­tion our objec­tions to Mr. Will’s silly col­umn in the spirit of true Chris­t­ian gift-​giving, espe­cially as we belat­edly pub­lish it on this Christ­mas morning.

There’s another small book – very thin and short – that both Mr. Will and Mr. Wald­fo­gel may wish to read. It takes an hour or two for most adults to com­plete, or a few hours more if read to a child (pre­sum­ing that the men like and/​or have chil­dren or grand­chil­dren). It’s called A Christ­mas Carol by Charles Dick­ens. Our read­ers may have heard of it, and it is avail­able free on-​line at many sites.

We strongly encour­age all our read­ers to read it every year so that they don’t end up like old Mar­ley, er, we mean Will.

God Bless us all, and we wish you a Merry Christ­mas and a Happy New Year.

–Andy, Jill and the gang

P.S. It’s late and there are prob­a­bly sev­eral typos. We will cor­rect them tomorrow.

There was a very inter­est­ing arti­cle in last Tuesday’s (Novem­ber 10) edi­tion of The Wall Street Jour­nal. It is enti­tled, It Just Isn’t Work­ing? Some File for Cus­tomer Divorce, and it relates how some small busi­nesses are elim­i­nat­ing prob­lem­atic customers.

In the paper edi­tion, the arti­cle appeared under the “Small Busi­ness’ ban­ner, and it now resides under a similarly-​labeled sec­tion of the web site; how­ever, it applies equally well to busi­nesses of all sizes.

We must admit, though, that the topic is espe­cially poignant for small busi­nesses because the decision-​maker and the imple­menter (exe­cu­tioner) are often one-​and-​the-​same per­son, and the cus­tomer may be a friend or acquain­tance. There’s no, “my boss told me that I have to do it” excuse when you are the boss. Note, how­ever, that many seemingly-​independent, entre­pre­neur­ial men still use their wives as rea­sons why some­thing can’t be bought or sold at a given price. We would never admit to such behav­ior, unless our Chair­man, Jill, per­mits us.

The main point of the arti­cle is that despite gen­er­ally tough eco­nomic times, a few small busi­nesses are find­ing that it is not worth­while to deal with cer­tain clients because those clients dis­pro­por­tion­ately con­sume time and resources given the rev­enue that they generate.

Dis­ci­pline! Discipline!

Though we sound like Colonel Hathi from The Jun­gle Book, there are a cou­ple of obvi­ous ways to min­i­mize the prob­a­bil­ity of enabling such problems:

1. Don’t appear too des­per­ate or too needy for the busi­ness. (In that sense, it is a lot like dating.)
2. Eval­u­ate the client, and have a slid­ing scale of prices rang­ing from lower prices for easy clients to higher prices for pains-​in-​the-​butt or annoy­ing ones, and stick to it.

Admit­tedly, such tac­tics may not be fea­si­ble with all prod­ucts and ser­vices. They gen­er­ally work bet­ter for ser­vices than prod­ucts, and for cus­tomized ser­vices, rather than uni­form or generic ones, and for short-​term projects rather than long-​term ones. How­ever, unless there is a sub­stan­tial ben­e­fit that can be derived from either exist­ing or prospec­tive clients because of the rela­tion­ship with a tire­some one, the tire­some one should pay for the irri­ta­tion, aggra­va­tion, dis­com­fort, and effort that they cause. (We’ll qual­ify this a bit below.)

If such cus­tomers are not will­ing to pay, then the sup­plier or ven­dor is, in fact, sub­si­diz­ing the oppor­tunis­tic behav­ior. More­over, if the sup­plier is not will­ing to sever the rela­tion­ship (or take actions that will change the other party’s behav­ior and prof­itabil­ity), then we humbly rec­om­mend that they not com­plain to every­one around them (thereby mak­ing their employ­ees’ and fam­i­lies’ lives more mis­er­able than need be).

That may seem a bit harsh, but it is con­sis­tent with our response when some­one com­plains about a spouse: divorce them, kill them, or shut-​up. Revealed pref­er­ence says that despite the com­plaints, if the ven­dor is unwill­ing to sever the rela­tion­ship, then he or she must find the arrange­ment to be accept­able. Some­times it is nec­es­sary to do unpleas­ant things – not immoral, uneth­i­cal, or ille­gal, just unpleas­ant ones – to get what one wants out of life.

So, be a good stoic, and live with it in silence (or turn it into a blog post).

We pre­fer to main­tain pric­ing dis­ci­pline because at the mar­gin, one’s will­ing­ness to lower the price or fee is often viewed as a sign of weak­ness. So, clients who make that infer­ence then often request more time and/​or resources. That means that a sign of weak­ness harms both rev­enue and resource-​usage, espe­cially for ser­vices that are pro­vided over time. (Note that this prob­lem can be mit­i­gated some­what by offer­ing a lower-​quality or less-​robust sub­sti­tute when­ever some­one wants a lower price, i.e., “we’ll lower the price, but you’re not getting…”)

By the way, another option is to write very, very detailed con­tracts, but those prepa­ra­tion costs only add to the resource con­sumed by such cus­tomers, and cov­er­ing those incre­men­tal, transaction-​related costs increase the nec­es­sary price to make the trans­ac­tion “worth­while.” (You can see why many folks open fran­chises where such devel­op­ment and prepa­ra­tion costs can be shared nation­ally or even globally.)

Two Costs to Consider

Before fir­ing a cus­tomer who is thought to be “unprof­itable” be sure to con­sider a few fac­tors to deter­mine if the cus­tomer is truly unworthy.

Clearly the choice of fir­ing or retain­ing a cus­tomer is a deci­sion, or the selec­tion (and imple­men­ta­tion) of one pos­si­ble alter­na­tive from sev­eral avail­able alter­na­tives. (Note that “do noth­ing” is often an avail­able course of action.) So, rather than be con­cerned with the report­ing of finan­cial results, or account­ing, per se, we are inter­ested in the costs and ben­e­fits that vary across alter­na­tives, the rel­e­vant costs and ben­e­fits.

In par­tic­u­lar, we are inter­ested in

1. Expected Rel­e­vant Ben­e­fits and Costs
2. Expected Oppor­tu­nity Costs

Note that these costs may vary through time as cir­cum­stances and capac­ity uti­liza­tion change. Tech­ni­cally, oppor­tu­nity costs, which we’ll define below, can be cat­e­go­rized as rel­e­vant costs, but we think that they are wor­thy of their own place on the list.

Read the rest of this entry »

The High Cost of Cen­tral­iza­tion at the Defense Department

Penny-​wisdom and Pound-​foolishness

A few weeks ago, we had a con­ver­sa­tion with the CEO of a mid-​sized orga­ni­za­tion who asked, “Why would I ever want my divi­sions to com­pete with one and other?”

It was posed as a rhetor­i­cal ques­tion, but when we fin­ished the “short ver­sion” of our answer – about 30 min­utes later – we were sure that he under­stood that, indeed, there are times when intra-​organizational com­pe­ti­tion is a good idea – by good we mean value– or wealth– or welfare-​maximizing. (Obvi­ously, it’s not all of the time, but there are times, and that’s part of what this post is about.)

Like trans­fer pric­ing, such com­pe­ti­tion can – actu­ally should – lead to dis­agree­ment and con­flicts within the orga­ni­za­tion (and in many cases dupli­ca­tion of effort and a seem­ing waste of resourse), but such poli­cies may still be opti­mal when they cre­ate cost-​effective dis­ci­plin­ing and con­trol mech­a­nisms. In other words, incur­ring mar­gin­ally higher costs is ben­e­fi­cial if the costly activ­i­ties gen­er­ate greater mar­ginal rev­enue or greater sav­ings elsewhere.

What’s It Have to Do with the DoD?

For­mer Sec­re­tary of the Navy John Lehman had a very inter­est­ing essay in the July 18^th edi­tion of The Wall Street Jour­nal: Waste­ful Defense Spend­ing Is a Clear and Present Dan­ger. In the essay, he com­plained about cost over­runs and the cur­rent, dis­grace­ful state of the bloated Depart­ment of Defense bureau­cracy. (But the gov­ern­ment will man­age health care effi­ciently, won’t it? Yeah, right.)

Based upon our read­ing of his essay, we think that the for­mer sec­re­tary missed the main cause of the inef­fi­cien­cies and esca­lated spend­ing that he crit­i­cizes.¹ Mr.Lehman men­tioned a few prob­lems and many symp­toms but not the root cause of the exor­bi­tant costs, which result from mis­guided and incom­pe­tent attempts at cost sav­ings. Yes, the cost over­runs result from poorly-​conceived attempts at cost sav­ings, espe­cially what we see as the over-​centralization of pro­cure­ment. (Why, it is almost like one of those unin­tended con­se­quences thin­gies that seems to sur­prise our elected offi­cials every once and awhile, and that one occa­sion­ally hears about in the news.)

We’ll explain our rea­son­ing below – this a very long post – but first we want to men­tion an aside related to fixed, sunk, and mar­ginal costs, and that’s – what should be – the well-​known dis­tinc­tion between aver­age total costs and mar­ginal costs. The dis­tinc­tion is quite well-​understood – or should be by any­one whose taken an intro­duc­tory micro­eco­nom­ics course and/​or under­stands divi­sion at a fourth-​grade level – but many folks, espe­cially our pub­lic ser­vants, still seem to get it wrong.

A Long Aside

Many, if not most, new defense pro­grams – e.g., new types of fight­ers, bombers, ships – involve huge, upfront design costs that are fixed with respect to the num­ber of units pro­duced. More­over, besides being fixed with respect to vol­ume, they also sunk or already incurred by the time that pro­duc­tion begins.

Now, for the moment we’re ignor­ing the pro­duc­tion change orders that Mr. Lehman noted are prob­lem­atic. How­ever, exclud­ing those ad hoc demands for change, once the design is final­ized, actual man­u­fac­tur­ing costs per unit tend to be set and are often a sur­pris­ingly small per­cent­age of total costs.²

So, once the design is set, the mar­ginal cost of each addi­tional unit is rel­a­tively small.³ How­ever, mar­ginal pro­duc­tion costs will depend upon on the quan­ti­ties ordered and the time­li­ness of the order (so that pro­duc­tion can be planned effi­ciently). One can think of this in terms of clas­sic labor/​capital cost-​volume-​profit analysis.

As we under­stand it, the mar­ginal costs – if known by any­one in the Pen­ta­gon or the gov­ern­ment – are rarely, if ever, divulged. If any per-​unit cost is reported, it is usu­ally a guess at the total aver­age cost, which is the sum of the actual mar­ginal cost and the aver­age fixed cost (plus the per­mit­ted per-​unit profit). The aver­age fixed cost is sim­ply total fixed costs divided by the num­ber of units produced.

So, as planned out­put is cut and then pos­si­bly fur­ther reduced (due to, say, high aver­age reported costs per unit) the aver­age fixed cost and, there­fore, the total cost per unit increases (at an increas­ing rate) with lit­tle true over­all cost sav­ings. We see that as the dan­ger of not learn­ing about frac­tions and divi­sion in the fourth grade, i.e., when the denom­i­na­tor decreases, the absolute value of the frac­tion increases.

Note: in com­pet­i­tive or semi-​competitive mar­kets, when a firm attempts to set prices based upon aver­age total costs – and the firm is not the most effi­cient pro­ducer or doesn’t dif­fer­en­ti­ate its prod­ucts in a suf­fi­ciently appeal­ing way – it may enable a down­ward (death) spi­ral of its own demand that even­tu­ally bank­rupts it. (Any­one heard of GM?) That is, when­ever the firm increases prices to “recover costs” fewer units are sold, and a smaller por­tion of fixed costs are cov­ered by sales; so, aver­age costs increase. Repeat­ing the ini­tial jus­ti­fi­ca­tion that the sales price must be high enough to cover fully-​reported unit costs, the firm raises prices, even fewer units are demanded, aver­age cost(s) increase, and the cycle repeats, ad infini­tum or, more pre­cisely, ad nau­seum until it dies or is bailed-​out.

We think that the above aside explains some of the higher unit costs and infla­tion that Mr. Lehman men­tions, i.e., the appli­ca­tion of a flawed heuris­tic or rule-​of-​thumb. How­ever, we think that the main causes of the increase in costs and losses in effi­ciency result are more struc­tural and relate to: (1) the rel­a­tively recent demand for increased cen­tral­iza­tion of pro­cure­ment and (2) the coin­ci­dent desire for (erst­while) stan­dard­iza­tion within the Depart­ment of Defense.

When Stan­dard­iza­tion Isn’t Cheaper

We’ll address this sec­ond cause first because we think that the goal of plat­form and asset stan­dard­iza­tion is a com­po­nent of and informs upon the larger costs asso­ci­ated with the over-​centralization of pro­cure­ment, but before con­tin­u­ing, it’s impor­tant to note that we’re being highly spec­u­la­tive in our con­jec­tures and hypothe­ses, but that’s our nature and a large part of – what we see as – our ever-​growing charm.

One would expect an eco­nomic – mean­ing an effi­cient – pro­cure­ment pro­gram involv­ing many assets and goods across many users would include both cus­tom and stan­darized designs. By def­i­n­i­tion, a pro­gram that is biased either way – too lit­tle or too much stan­dard­iza­tion – would lead to higher costs for a given level of per­for­mance and/​or worse per­for­mance for a given level of cost than could oth­er­wise be obtained.

Two Ways to Stan­dard­ize Designs

When attempt­ing to design com­mon or stan­dard­ized prod­ucts to suit a vari­ety of users, we think there are two mian and oppos­ing philoso­phies, which we’ll refer to as (1) the inter­sec­tion approach and (2) the union approach.

1. Inter­sec­tion approach makes the item as sim­ple as pos­si­ble by pro­vid­ing only the shared func­tion­al­i­ties among a set of poten­tial cus­tomers and uses, like, say, a double-​edged knife – one ser­rated and one not for every­one who wants both capa­bil­i­ties. This is approach is most likely to gen­er­ate design and pro­duc­tion sav­ing – at the expense of reduced functionality.
2. Union approach attempts to design the item to be all things to all users – kind of like a Swiss Army knife – that incor­po­rates all of the var­i­ous desired char­ac­ter­is­tics that var­i­ous users may have been demanded.⁴ This approach is likely to have much higher design and man­u­fac­tur­ing costs but pro­duce more use­ful­ness. How­ever, it may also cause cer­tain users to incur higher filed (usage) costs.

Both the inter­sec­tion and the union approaches involve mak­ing trade-​offs dur­ing the design stage. Obvi­ously, those trade-​offs involve both costs and func­tion­al­ity in the var­i­ous fields of use. We’ll cat­e­go­rize the costs, includ­ing the oppor­tu­nity costs, into five cat­e­gories although we do under­stand that there are other equally valid ways to clas­sify them, e.g., design, pro­duc­tion, usage.

Five Costs Asso­ci­ated with Stan­dard­ized Designs

There are many poten­tial costs asso­ci­ated with uneco­nom­i­cal stan­dard­iza­tion:: (1) the increased time, effort, and cost of design (when, for exam­ple, con­trac­tors use the union approach an attempt to cram every service’s unique spec­i­fi­ca­tions into a sin­gle, fin­ished prod­uct); (2) the increased care and cost of pro­duc­ing to tighter tol­er­ances that is often required when prod­ucts become more com­plex and com­pli­cated (some­times more than they should be); (3) the costs of spe­cial requests and pro­duc­tion changes because the com­pro­mised near-​final prod­uct meets no sin­gle consumer’s unique needs; (4) com­pro­mised oper­a­tion and higher over­all oper­at­ing costs because the final prod­uct may be more or less than required, espe­cially when only spe­cific, lim­ited –but dis­tinct – func­tion­al­ity is demanded; and (5) higher fail­ure costs – both repair costs and failure-​related costs. These costs need not be finan­cial. The fail­ure of equip­ment in bat­tle can lead to loss of life and tac­ti­cal or strate­gic defeat; so, we’d rather each ser­vice use safer rather than sorry designs.

1. Lengthy & Expen­sive Design Processes

Because the process involves at least one more step, determing com­mon requests under the inter­sec­tion approach would lead to more lengthy design times and more expen­sive processes than if each ser­vice acted and pur­chased in a com­pletely autonomous manner.

How­ever, these cost dif­fer­en­tials would be sub­stan­tially greater under the union approach. As more and var­ied (and pos­si­bly con­flict­ing) spec­i­fi­ca­tions are demanded under the union approach, more design effort, time and cost are incurred. That is espe­cially true if the desired spec­i­fi­ca­tions con­flict with each other. For exam­ple, if the asset or good is sup­posed to be both light and strong – like a durable note­book com­puter or an easy-​to-​carry M4 car­bine that is light but has a suf­fi­ciently thick and heavy bar­rel to with­stand the intense heat of rapid and repeated fire. Obvi­ously, that’s true within one ser­vice, but the issue is exac­er­bated – per­haps accel­er­ated is a bet­ter word – when addi­tional, con­flict­ing spec­i­fi­ca­tions are demanded across users.

Under the union approach, not only are ini­tial designs of more com­plex prod­ucts more expen­sive to com­plete, but due to the inher­ent com­plex­ity and inter-​relationships, revi­sions are, too. Unless there is com­plete mod­u­lar­ity – like swap­ping out a computer’s SATA hard drive or a USB mouse or VGA mon­i­tor – revi­sions must con­sider those inter-​relationships and effects on the other desired prop­er­ties and func­tion­al­i­ties as design char­ac­ter­is­tics are changed.⁵

Mr. Lehman notes that “there has been a loss of dis­ci­pline and con­trol over equip­ment require­ments and a surge in gold-​plating in all Pen­ta­gon pro­grams.” We view this as an impli­ca­tion of the “union” approach of attempt­ing to be all things to all armed ser­vices. With many con­flict­ing require­ments, the “gold-​plating” occurs because it is often nec­es­sary to use state-​of-​the-​art designs, mate­ri­als, and pro­duc­tion processes (to pro­duce assets with many con­flict­ing design requirements).

In sum, and very gen­er­ally, one can think of (1) the time to com­plete a design and (2) the cost of both the design process and pro­duc­tion to be increas­ing and con­vex func­tions of the num­ber and strin­gency of dif­fer­ent per­for­mance attrib­utes and func­tion­al­i­ties. Changes in tech­nol­ogy would shift and rotate those rela­tion­ships, but that’s a story for another day.)

2. Increased Care & Costs of Production

Under the union approach, for a given level of tech­nol­ogy, sat­is­fy­ing mul­ti­ple and con­flict­ing require­ments gen­er­ally requires the pro­ducer to incur higher fail­ure “pre­ven­tion” costs, which are the upfront costs of increas­ing the prob­a­bil­ity of con­form­ing out­put, and higher process mon­i­tor­ing costs, i.e., more care­ful and super­vised pro­duc­tion. Tech­no­log­i­cal inno­va­tion, by upwardly shift­ing per­for­mance attrib­utes and pro­duc­tion functions, can mit­i­gate these costs; how­ever, per our point in (1) above, often the com­plex­ity of design does not per­mit such easy, implication-​free substitutions.

It seems that much of the lay­ered bureau­cracy within defense con­trac­tors and the Depart­ment of Defense involves record­ing and cer­ti­fy­ing and audit­ing many of these quality-​related process vari­ables. Iron­i­cally, it is often in the name of increas­ing effi­cien­cies and cost man­age­ment. What a joke!

3. Costs of Spe­cial Requests & Design Modifications

Under either approach, when design­ing a “stan­dard­ized prod­uct,” there seems to be a rea­son­able chance that no one will be com­pletely sat­is­fied with the final design trade-​offs. Under the inter­sec­tion approach, that’s because each user may be denied a unique and impor­tant func­tion­al­ity, whereas under the union approach, it is not nec­es­sar­ily fea­si­ble to pro­duce a union of all demanded char­ac­ter­is­tic, or the inclu­sion of all demanded func­tion­al­i­ties may require some users to bear higher costs than they oth­er­wise would have done so.⁶ As such, the final design may be a prover­bial jack-​of-​all-​trades, but mas­ter of none.

So, one can eas­ily imag­ine ad hoc demands for changed spec­i­fi­ca­tions dur­ing or imme­di­ately prior to the start of pro­duc­tion runs for par­tic­u­larly branches of the ser­vice. For exam­ple, when a stan­dard item’s paint is changed from Army green to Air Force blue, one can imag­ine the Air Force other changes to the item, and those requests increase pro­duc­tion time and costs. Thus, the promised sav­ings from stan­dard­ized design and pro­duc­tion are likely to be reduced if not eliminated.

We sus­pect that given the dif­fer­ent ser­vices and the real dif­fer­ences in their needs, there is far less real­ized stan­dard­iza­tion than promised or ini­tially hoped for under a “joint” or union approach. So rather than sep­a­rate, dupli­cate, upfront designs, there are last-​minute, dupli­cate, ad hoc designs. We’d argue that such ad hoc changes and designs are more expen­sive than planned ones.

In fact, Mr. Lehman men­tions some­thing about 75 change orders per week for (some set or sub­set) of defense con­tracts. We’re not sure what that num­ber should be com­pared against or what his­tor­i­cal lev­els were, but we’d hypoth­e­size that many of those requests are adap­ta­tions of a sup­pos­edly stan­dard­ized, joint design to the dif­fer­ent demands of the Army, Navy, Marines, etc.

4. Com­pro­mised Use & Higher Oper­at­ing Costs

Under the inter­sec­tion approach, this com­pro­mised use would include the cumu­la­tive oppor­tu­nity costs of each user not hav­ing the func­tion­al­ity that they desire (and need). Unfor­tu­nately, this is one area where the non-​financial costs of com­pro­mised usage can be enor­mous. They include the tragic loss of sons and daugh­ters, moth­ers and fathers, and broth­ers and sis­ters. They may also include the loss of tac­ti­cal goals (bat­tles) and strate­gic opb­jec­tives (wars).

Those same costs can arise under the union approach, too. That design approach can cre­ate (the metaphor­i­cal) “too much bag­gage.” By pro­vid­ing func­tion­al­ity for one par­tic­u­lar user in a shared design, it is pos­si­ble to reduce it for another by over­bur­den­ing them. For exam­ple, mak­ing some­thing, say, rugged for one group may make it too heavy for another or mak­ing it light for one may make it too frag­ile for another.

Think of a firm that pro­vides each employee with a nearly-​indestructible, ruggedi­zed, $4,000 lap­tops, with say, a 13-​inch screen. Many employ­ees would be better-​served with desk­top, a cheaper note­book, a tablet, a very cheap net­book, a hand­held device, or sim­ply a dumb cell phone. For exam­ple, we’d imag­ine that there are pilots in the mil­i­tary who are assigned planes that can’t fly slow enough to opti­mally per­form their mis­sions, and we’re sure that some­one with mil­i­tary expe­ri­ence could pro­vide more examples.

5. Higher Fail­ure & Repair Costs

Related to com­pro­mized usage and higher oper­at­ing costs are higher failure-​related costs, includ­ing – but not lim­ited – to repair costs.

All things equal, the Swiss Army approach of more com­plex­ity and more parts would gen­er­ally lead to higher exter­nal (field) fail­ure rates and higher repair and replace­ment costs. This is espe­cially prob­lem­atic when fail­ure or one, oth­er­wise extra­ne­ous or unnec­es­sary com­po­nent, pre­vents oper­a­tion of the unit as a whole.

Often these higher costs induce those respon­si­ble for the asset to use it less than it should or could be used, e.g., think of the beau­ti­ful and expen­sive “good” china and crys­tal that sits in the cab­i­net rather than on the table when there are no guests.⁷

As with com­pro­mised func­tion­al­ity, besides the incur­rence of higher finan­cial costs, the occur­rence of increased exter­nal fail­ures and break­downs (and pos­si­bly lengthy repair processes and dura­tions) leaves the nation less pre­pared to defend itself and, most impor­tantly, may expose our sol­diers, sailors, and marines to unnec­es­sary and lethal dangers.

Given these five poten­tial cost cat­e­gories, one should ques­tion whether the devel­op­ment of “joint” pro­grams is either effi­cient or effec­tive. We doubt it. In our mind, a proper account­ing would show few cost sav­ings, many delays, and many oppor­tu­ni­ties for errors, mis­takes, and failures.

Under either approach, in our mind, unnec­es­sary stan­dard­iza­tion is a symp­tom of the big­ger prob­lem: an uneco­nom­i­cal centralization.

The Fail­ure of Cen­tral­ized Procurement

Admit­tedly, it is dif­fi­cult to com­pare the cost and waste of the cur­rent, rather cen­tral­ized sys­tem with the costs (and ben­e­fits) under a more decen­tral­ized pro­cure­ment struc­ture because there is only one Pen­ta­gon, and we can’t observe the per­for­mance under the alter­na­tive struc­ture. How­ever, we argue that beyond too much stan­dard­iza­tion, too much cen­tral­iza­tion can lead to higher costs, reduced effi­ciency, and reduced effec­tive­ness and per­for­mance in the field, and that is the cur­rent state of defense pro­cure­ment in the USA.

Sec­re­tary Lehman – and oth­ers – pro­vide evi­dence that the cur­rent, rather-​centralized, process is bro­ken. In our mind, no stronger evi­dence exists than the recently signed law that will add 20,000 more bureau­crats to “reform” defense acqui­si­tions. Unfor­tu­nately, that will likely lead to an even more cen­tral­ized struc­ture.⁸ That men­tal­ity is very much like what we see with health care, where the fail­ure of the Med­ic­aid and Medicare to con­trol costs has led to the call from some par­ties for more cen­tral­ized resource allocation.

How to Think about the Problem

At the bot­tom ofOur Con­trol Frame­work page, we list the pos­si­ble ben­e­fits and costs of decen­tral­iza­tion – i.e., del­e­ga­tion and auton­omy – ver­sus a more cen­tral­ized structure.

The costs asso­ci­ated with auton­omy relate to (1) self­ish­ness and (2) igno­rance, and include those asso­ci­ated with the dupli­ca­tion of effort. Such dupli­ca­tion can occur because one sub­or­di­nate doesn’t know that another is per­form­ing the same activ­ity (igno­rance) or because one of the par­ties wants to “do it my way” (selfishness).

While such dupli­ca­tion may seem bad on the sur­face, elim­i­nat­ing it and the asso­ci­ated costs need not be opti­mal – in either the short-​term or the long-​term. That’s because elim­i­nat­ing the costs of decen­tral­iza­tion usu­ally destroys the asso­ci­ated ben­e­fits, too, and those ben­e­fits may be greater than the costs. Thus, the desire to elim­i­nate gov­ern­ment waste may be very myopic behav­ior and is not much dif­fer­ent than firms that attempt to max­i­mize prof­its on a quarter-​for-​quarter basis rather than their long-​term value.⁹ Such “waste” may be a byprod­uct of the most effi­cient struc­ture, but one wouldn’t know that if their response were noth­ing more than a thought­less, knee-​reaction to elim­i­nate it.

We sus­pect that the demand for more cen­tral­ized struc­tures within the Defense Depart­ment arose from such reac­tions – reported inci­dents that showed such dupli­ca­tion and per­ceived “waste.” For exam­ple, some­one may have observed that the Army was sep­a­rately doing the same thing as the Navy was doing, and rhetor­i­cally asked, “why can’t we com­bine them?” and thus, oper­a­tions were “stream-​lined” before the rhetor­i­cal ques­tion could be intel­li­gently answered.

Given that there is not a goal of profit-​maximization at the Defense Depart­ment, we’d hope that its goal involves max­i­miz­ing the nation’s defense capa­bil­i­ties – some­thing like keep­ing our coun­try and its cit­i­zens safe sub­ject to var­i­ous uncon­trol­lable envi­ron­men­tal and bud­getary con­straints. That’s a much more neb­u­lous goal than mere profit max­i­miza­tion, and it makes study­ing (and man­ag­ing) gov­ern­ments (and other types of not-​for-​profit orga­ni­za­tions) more chal­leng­ing than ana­lyz­ing and man­ag­ing firms. Regard­less, one can still man­age thought­fully if one chooses, but our point is that the knee-​jerk reac­tions of Con­gress­men and bureau­crats to con­sol­i­date pro­grams and cen­trally admin­is­ter them is no dif­fer­ent than other myopic, inef­fi­cient, cost-​minimizing behav­ior. In other words, try­ing to min­i­mize a sub­set of costs need not max­i­mize value, effi­ciency, effec­tive­ness nor even min­i­mize total costs. We think it requires a par­tic­u­larly high degree of (self) dis­plicine to incur these costs because they related to oth­ers’ behav­ior. (See our essays, Com­mon Man­age­r­ial Mis­takes in Decen­tral­ized Orga­ni­za­tions and Strate­gic Con­sis­tency and Man­age­r­ial Dis­ci­pline for more on this topic.)

Again, per the mar­ginal ben­e­fits of decen­tral­iza­tion that are listed on the bot­tom of Our Con­trol Frame­work page, it is easy to believe that:

1. Sub­or­di­nates (on land, in the sea, and in the air) are bet­ter informed…

… about the cir­cum­stances and envi­ron­ments that they face or may face and about their own needs. Thus, they are bet­ter able to pre­cisely spec­ify the assets, goods, and mate­ri­als that they need to accom­plish their spe­cific missions.

2. Subordinates have more exper­tise and knowledge…

…to pro­cure what they need to adapt their capa­bil­i­ties to their cur­rent and prospec­tive envi­ron­ments, includ­ing likely foes and loca­tions. They have more exper­tise about how they fight and train and use the assets under con­sid­er­a­tion (and what would be opti­mal con­fig­u­ra­tions of those assets).¹⁰

3. Sub­or­di­nates can respond quicker and decide faster…

…than (a mono­lithic) cen­tral­ized author­i­tyor author­i­ties like the um, er, the Pen­ta­gon and the DoD and the Con­gres­sional bureau­cracy. The inge­nu­ity and cre­ativ­ity that Amer­i­cans so love, espe­cially in their armed forces, is crushed by such processes. Thus, one sees it in the change orders and field (and bat­tle­field) adap­ta­tions because they were given com­pro­mised assets and equip­ment – rather than spe­cial­ized equip­ment. Why not per­mit (or induce) them to be ingen­u­ous in the design phase to meet their par­tic­u­lar needs – espe­cially when their lives are at stake?

The elon­gated design times would likely be sub­stan­tially reduced if the branches of the ser­vice were given more auton­omy to pur­chase to their spe­cific needs. It would take less time to accu­mu­late demands and con­sider the trade-​offs, and the com­pro­mise on those demands (across the ser­vices) would be elim­i­nated. That could per­mit shorter asset life cycles and the ear­lier imple­men­ta­tion of enhanced func­tion­al­i­ties and tech­nolo­gies of later gen­er­a­tion platforms.

4. Sub­or­di­nates may be bet­ter motivated…

and may bet­ter under­stand the rela­tion­ship of their actions to their organization’s goals and envi­ron­ment. Per Mr. Lehman’s essay, “there has been an oblit­er­a­tion of clear lines of author­ity for man­ag­ing pro­cure­ment pro­grams.” In other words, your service’s piece is small, it’s not in your bud­get, the joint-​platform has to be com­pleted and won’t be dis­carded, and you can always try to change it later when they’re pro­duc­ing your units. So why bother fuss­ing with con­trol­ing the design time or costs? What can one per­son do any­way, and who can blame them free­load­ing?¹¹

Can the reader imag­ine a burea­crat step­ping up and tak­ing respon­si­bil­ity for a project and its suc­cess? (TARP anyone?)

The branches of the ser­vice would also be bet­ter moti­vated to pro­cure effi­ciently if they (a) faced stronger, inter-​service com­pe­ti­tion for pro­cure­ment bud­get dol­lars and (b) per­haps some types of dead­lines, and that com­pe­ti­tion would con­trol costs and the gold-​plating Mr. Lehman men­tions. The for­mer, (a), is con­sis­tent with our obser­va­tion that not all intra-​organizational com­pe­ti­tion is bad, and (admit­tedly) the lat­ter, (b), is some­thing that we need to think more about. (In many ways, these elon­gated, decen­tral­ized design processes remind us of the eight-​year-​old hole in the ground in Lower Man­hat­tan. Fol­low­ing a WSJ edi­to­r­ial on July 21, we wrote about it in The Right Com­par­i­son and the related Cor­po­rate Projects and D-​Day.)

5. Infor­ma­tion sys­tem costs may be lower…

…as less data col­lec­tion and aggre­ga­tion is necessary.

We real­ize that we’re being unre­al­is­tic in hop­ing that in the future, engi­neers at con­trac­tors may be per­mit­ted to account for their time in greater than six minute incre­ments.¹² Such poli­cies are stu­pid (self-​defeating), demor­al­iz­ing, and dehu­man­iz­ing, but we doubt that such reg­u­la­tions will be relaxed. How­ever, in the con­text of this essay, less cen­tral­ized data col­lec­tion and aggre­ga­tion means fewer par­a­sitic bureau­crats who are aggre­gat­ing already aggre­gated reports for no other pur­pose than to aggre­gate and bureau­cra­tize. For that rea­son, alone, cit­i­zens should pre­fer a more decen­tral­ized approach to procurement.

The real­iza­tion of many of these ben­e­fits would indi­vid­u­ally (and col­lec­tively) reduce the five costs of excess stan­dard­iza­tion that we men­tioned above; how­ever, they are broader and involve more than just stan­dard­iza­tion of par­tic­u­lar assets and goods, i.e., the joint plat­forms as they involve over­all costs sav­ings and improved effi­ciency and effectiveness.

We’re not under­stat­ing the dif­fi­culty of solv­ing such a cru­cial, yet amor­phous, prob­lem, espe­cially given the num­ber com­pet­ing and con­flict­ing claimants – both within and out­side of the gov­ern­ment. We are sim­ply argu­ing that cen­tral­iza­tion of resource allo­ca­tion need not be opti­mal, and at the present time it is likely sub-​optimal within the Depart­ment of Defense.

Finally, we see no rea­son to believe that the cen­tral­iza­tion of health care pro­cure­ment will be any more effec­tive than the failed cen­tral­iza­tion of defense pro­cure­ment (and already failed health­care pro­grams). In nei­ther case, should one expect that addi­tional cen­tral­iza­tion will com­pen­sate for the wastes asso­ci­ated with excess centralization.

P.S. We’ll likely edit this in the near future to elim­i­nate typos and add skipped words and improve the sen­tence struc­ture of cer­tain paragraphs.

Copy­right © 2009 Spero Consulting.

---

Foot­notes:

There’s an excel­lent opin­ion col­umn in yesterday’s (May 28) edi­tion of The Wall Street Jour­nal. It is Crazy Com­pen­sa­tion and the Cri­sis by Alan S. Blinder.

Why do we write that it is “excel­lent” the dear reader may ask?

Well, for the obvi­ous (and self-​serving) rea­son that we have been writ­ing the same cri­tiques on these pages for much of the past year or so.

Mr. Blinder iden­ti­fies sev­eral prob­lems that cre­ated the poten­tial for the cri­sis and its sub­se­quent real­iza­tion.¹ We will cat­e­go­rize the prob­lems that he iden­ti­fies as:

1. Wrong legal form/​organization struc­ture for some firms,
2. Incom­pe­tent boards, and
3. Lax con­trols and poorly-​designed incentives.

He treats them in a dif­fer­ent order than we list them; we’re going from top-​to-​bottom, which is con­sis­tent with Our Con­trol Frame­work. Clearly, the three cat­e­gories are related. For exam­ple, see our pop­u­lar post, SOX’s Roles in the Finan­cial Cri­sis of ‘08, which hits on all three top­ics, and crit­i­cizes gov­ern­ment reg­u­la­tion to boot. In our mind, they all pro­vide evi­dence of the fallen nature of man. (We’re not com­plain­ing about that nature. We accept it in our­self and, to a lesser extent, in oth­ers. We’re only try­ing to profit from it.)

Wrong Legal Form/​Organization Structure

We wrote about this on Sep­tem­ber 26, 2008, when we asked Will Invest­ment Banks Go the Way of the Dinosaur? In that post we spec­u­lated that part­ner­ships may make a come­back because “They pro­vide con­trol mech­a­nisms and lev­els of over­sight and scrutiny that seem dif­fi­cult to dupli­cate in pub­lic corporations.”

Mr. Blinder made explicit what was implicit in our post: the dif­fer­ence between one’s level of risk-​taking when man­ag­ing OPM (Other People’s Money) ver­sus what he refers to as MOM (My Own Money), or one’s own money.² Those fac­ing unlim­ited per­sonal losses tend to be more con­ser­v­a­tive than those with lim­ited losses.

In Jan­u­ary, in a cri­tique of The Wall Street Jour­nal’s edi­to­r­ial board, What Did They Expect?, we wrote, “We also dis­agree with their [the edi­to­r­ial board’s] assess­ment that “com­pen­sa­tion lev­els are a busi­ness judg­ment made under the pres­sure of com­pe­ti­tion.” That might be true if the firms were part­ner­ships or oth­er­wise privately-​owned, there was no agency costs, and there was no self-​dealing, i.e., the firms were run by inde­pen­dent and knowl­edge­able boards.”

But with D & O (direc­tors’ and offi­cers’) insur­ance, the lim­ited down­side of losses severely decom­presses that so-​called “pres­sure of com­pe­ti­tion” for boards. More­over, share­hold­ers of bank hold­ing com­pa­nies (and other cor­po­ra­tions, too) implic­itly per­mit­ted man­agers to take greater risks. In fact, Mr. Blinder seems unwill­ing to blame share­hold­ers when almost every stock­holder was quite capa­ble of sell­ing their stakes. So, we have no sym­pa­thy for folks who wanted the oppor­tu­nity for large gains with­out bear­ing poten­tial lia­bil­i­ties if the firm.³

Incom­pe­tent Boards

While “Incom­pe­tent Boards,” may seem a bit harsh to some, we think that it is milder than many alter­na­tive and equally fair char­ac­ter­i­za­tions, and there is no short­age of evi­dence. See Direc­tors Are Faulted at Home Loan Banks for example.

Reg­u­lar read­ers will note that we often ask whether a party is igno­rant or cyn­i­cal, and in this case we’d pre­fer to believe that many direc­tors were unqual­i­fied to under­stand the uncer­tain­ties and risks asso­ci­ated with invest­ing and trad­ing, par­tic­u­larly with deriv­a­tives and other struc­tured prod­ucts. In some way, that seems more “decent” and eth­i­cal than the alter­na­tive: the cyn­i­cal and devi­ous behav­ior of under­stand­ing the poten­tial for loss but ignor­ing it due to one’s own lim­ited lia­bil­ity.⁴

For exam­ple, with the recent changes in the com­po­si­tion its board, Citi­corp has as much as admit­ted the lack of req­ui­site exper­tise of its past board. We’ve writ­ten about these top­ics in the past, par­tic­u­larly in: The Fail­ure of Boards to Direct, The Seventy-​Year-​Old Teenager, When the Going Gets Tough…Quit, and Idio­syn­cratic and Con­cen­tra­tion Risk, Again. (Update: within hours of pub­lish­ing this post, B of A announced that one of its direc­tors was resign­ing: see BofA Says Sloan Quits Board Seat. There was much spec­u­la­tion that it was due to gov­ern­ment pressure.)

Those (gen­er­ally weak and) incom­pe­tent boards per­mit­ted senior man­agers to main­tain the lax con­trols and poorly-​designed incen­tives about which we have often writ­ten, and here is a summary.

Lax Con­trols and Poorly-​designed Incentives

As Mr. Blinder notes, poorly-​designed incen­tives – pri­mar­ily via com­pen­sa­tion schemes – led to ex post “exces­sive” risk-​taking. We write ex post as in 20 – 20 hind­sight as in “there are mas­sive losses, so some­one must have done some­thing wrong,” but, in fact, we’re note using that logic. Instead, we note that there was no short­age of indi­vid­u­als warn­ing about the risk and uncer­tain­ties ex ante.

Unfor­tu­nately, many such folks were dis­missed either fig­u­ra­tively or lit­er­ally by senior man­age­ments. (It’s anal­o­gous to the SEC’s treat­ment of Harry Markopo­los. See Cas­san­dra, the SEC and Mr. Mad­off.) More­over, it is con­sis­tent with the per­spec­tive that risk man­agers gen­er­ate no rev­enue and are costs to be min­i­mized (and often voices to be ignored).

So, yes, traders (and their man­agers) took gam­bles because they bore (or thought they bore) lim­ited down­side risk but instead focused on the poten­tial for sub­stan­tial (enor­mous) com­pen­sa­tion rewards, but lax con­trols and igno­rance are big­ger issues than just poorly-​designed com­pen­sa­tion schemes because said traders were allowed to take those gam­bles with OPM.

That lack of con­trol has many facets, but can be sum­ma­rized in terms of as greed, igno­rance, and inse­cu­rity. Notice that, of course, those emotions/​human con­di­tions are always present, but pre­cisely the job of senior man­agers (and boards and own­ers) to design schemes and mech­a­nisms that take those as given and mit­i­gate them – rather than exac­er­bate them – while the orga­ni­za­tion attempts to achieve its objec­tive. (We’ll have more to say about that below.)

Igno­rance, and its rel­a­tive, inse­cu­rity, were cru­cial to the con­trol fail­ures. Few folks are will­ing to admit that some­thing is immea­sur­able or nearly impos­si­ble to quan­tify because that can be turned-​around and used against them as a per­sonal short-​coming:, e.g., “that’s just because he doesn’t know enough.” So, per­sonal inse­cu­rity and incen­tives often induce employ­ees to “take the easy way out” and endorse or embrace a sim­plis­tic and inap­plic­a­ble val­u­a­tion or risk model. 

For exam­ple, in early Novem­ber, we wrote The Under­state­ment of the Year! in response to an arti­cle in The Wall Street Jour­nal enti­tled, Behind AIG’s Fall, Risk Mod­els Failed to Pass Real-​World Test. While the entire post is rel­e­vant to this dis­cus­sion, we par­tic­u­larly like this extended excerpt:

The prob­lem, dear reader, is that few senior man­agers (and almost no board members) understand the val­u­a­tion and risk mod­els used for secu­ri­ti­za­tions, and many of the traders, con­sul­tants, and ana­lysts who wield such tools often suf­fer from, what one may call, “fram­ing” issues; we don’t mean that aspect of home con­struc­tion despite its recent relevance.

We mean that if one’s only tool is a ham­mer, then lots of things look like nails. The metaphoric ham­mer may be an intan­gi­ble Visual Basic or “C” pro­gram­ming algo­rithm, but the point remains the same; it’s just harder for senior man­age­ment to see what one is pound­ing in their cubi­cle, office, or trading-​floor seat.

To be sure, if any­one within most of the larger firms would have com­plained of the sys­tem­atic risk — and how every­thing could go bad all at once — and the inap­plic­a­bil­ity of the stan­dard mod­els, which gen­er­ally don’t per­mit such events, then that per­son most cer­tainly would have been told that they don’t know what they’re talk­ing about. Pos­si­bly, that they are unso­phis­ti­cated or too negative.

Ear­lier this week in Uncer­tainty: In God We Trust, we noted “Too many senior man­agers neglected their respon­si­bil­i­ties and per­mit­ted the sub­sti­tu­tion of cal­cu­la­tions for thoughts.” That as been a pet peeve of ours for quite some time and is the antithe­sis of our motto: thought before cal­cu­la­tion. See The Dif­fer­ence Between Risk and Uncer­tainty for a rel­a­tively short expo­si­tion of the issues.

Those dys­func­tional behav­iors were not nec­es­sar­ily mali­cious or anti-​social by intent, but does that mat­ter, espe­cially since thought­ful design of con­trol mech­a­nisms could have inhib­ited them? See Prin­ci­ples Lost and More, in which we con­trast Saint Thomas More’s actions in the 16^th cen­tury with the more recent actions of many less holy indi­vid­u­als prior to and dur­ing the Finan­cial Cri­sis; there’s a rea­son he’s a Saint and we’re not.

We’ve writ­ten much, much more on this topic, but as we noted in The Prob­lem of Induc­tion, we’re not under­es­ti­mat­ing the dif­fi­culty of the prob­lems faced by traders, struc­tur­ers, and risk man­agers. In fact, if any­thing, we’re overly con­ser­v­a­tive by stat­ing that not all uncer­tain­ties and losses can be quan­ti­fied and the prob­lems are much more dif­fi­cult than some sup­pose and/​or communicate.

What To Do?

Unfor­tu­nately, Mr. Blinder notices that there has been little-​to-​no struc­tural change in cor­po­rate gov­er­nance. He attrib­utes the dif­fer­ences in mar­kets – the illiq­uid­ity or lack of trad­ing – to fear, rather than to newly designed or revised con­trols, and that seems about right to us. As we noted last month in Learn­ing the Dif­fer­ence Between Risk and Uncer­tainty, or not, job descrip­tions and hir­ing require­ments for many trad­ing and risk man­age­ment posi­tions don’t seem to have changed; so, it doesn’t seem the firms have “re-​engineered” or redesigned their oper­a­tions or controls.

In Octo­ber, we wrote a tongue-​in-​cheek post about The Role for Sur­vival­ists and Depres­sives in Uncer­tainty Man­age­ment, but in all seri­ous­ness, hir­ing such per­son­al­i­ties and lis­ten­ing to them is one way to com­pen­sate for flawed risk models.

To be fair, we have read about a few firms, like UBS, that have changed their com­pen­sa­tion schemes to include fea­tures like claw­backs. See Claw­backs: the Good, the Bad, and the Ugly and Incen­tives at UBS and in Gen­eral. How­ever, it is not clear whether such changes have been thought­fully man­aged. As we men­tioned in Busi­ness Schools, Incen­tives, Uncer­tainty, and the Finan­cial Cri­sis, it seems that lit­tle has been done because: (1) such incen­tive prob­lems are very chal­leng­ing to solve, and (2) uni­ver­si­ties don’t do a par­tic­u­larly good job of train­ing busi­ness stu­dents to solve them. (Of course, for the right fee, we would be glad to help.)

So what to do?

Mr. Blinder calls for change, but doesn’t exactly explain how or what.

We’ve made sev­eral rec­om­men­da­tions in past, includ­ing this post from early Octo­ber: Elim­i­nate Pro­pri­etary Trad­ing at Insured Insti­tu­tions. Every­thing in it – and there’s a lot – holds up well, and we’ve not heard a com­pelling argu­ment against such a ban. As we wrote back then:

We’re com­pletely for the free-​market—more so than most bank man­agers — but until such insti­tu­tions for­sake their gov­ern­ment insur­ance, we’ll insist that they have an oblig­a­tion to the cit­i­zenry — through the gov­ern­ment — to behave in a respon­si­ble, low risk man­ner. If that gen­er­ates lower returns for them on aver­age, then so be it. That’s the nature of the risk-​return spec­trum and their legal and fiduciary responsibilities…

We think that such a ban is fea­si­ble and would sub­stan­tially mit­i­gate many of the risks that those banks by elim­i­nat­ing the (socially) unde­sir­able behavior.

Now, that (max­i­mum) risk-​seeking behav­ior is not uni­ver­sally unde­sir­able, but it is within sub­si­dized insti­tu­tions. We’re all for per­mit­ting “prop” struc­tur­ers and traders to oper­ate in unreg­u­lated part­ner­ships and hedge funds, and wish such orga­ni­za­tions the best of luck.

P.S. Although this post is rife with links, we’ve writ­ten much, much more about the top­ics of risk man­age­ment, incen­tives, and the cri­sis. Feel free to peruse the archives, and let us know if we’re wrong about any­thing – other than a few predictions.

P.P.S. As posted, this is rather long, and we’ll likely revise it in the near future as we dis­cover typos, etc.

What Should It Mean to Earn a Master’s Degree?

We don’t answer that ques­tion here, but shouldn’t one be required to mas­ter something?

It Was a Mat­ter of Time

Since early Octo­ber, we’ve won­dered when we’d see the first edi­to­r­ial crit­i­ciz­ing MBAs and busi­ness schools for their role in the ongo­ing finan­cial cri­sis.¹ In our mind, much of the blame should be shared between busi­ness types, i.e., MBAs, and so-​called “quants,” with the major­ity of the blame placed on senior man­agers who per­mit­ted lax con­trols and mis­aligned incen­tives to exist.

We didn’t write about it when the thought orig­i­nally occurred to us nor dur­ing the inter­ven­ing six months-​or-​so, but we’ve been tempted to write on any num­ber of occasions.

Two events occurred last week that moti­vated us to write today. First, our excel­lent, for­mer TA, Brid­get Ardoyno, wrote to us that she has been blog­ging at http://​econ​mom​.blogspot​.com, and that reminded us of teach­ing MBAs (but in a good way).

The Main Shortcoming

The other event was the appear­ance of an excel­lent opin­ion col­umn, How Busi­ness Schools Have Failed Busi­ness, in last Friday’s edi­tion of The Wall Street Jour­nal. The col­umn, by Michael Jacobs, lists three main fail­ings of busi­ness schools with respect to the teach­ing and the cri­sis, but in fact, his three are all exam­ples of the lack of the qual­ity instruc­tion regard­ing con­trol and incen­tives.² Basi­cally, incen­tive issues are a type of con­trol prob­lem that arise in decen­tral­ized orga­ni­za­tion, where sub­or­di­nates are per­mit­ted a degree of auton­omy to act as they see fit.

The Root Causes

There is much to like about Mr. Jacobs’s crit­i­cism of busi­ness schools. How­ever, while we real­ize that edi­to­r­ial space is limited, he ignores the two main causes of the prob­lems that he iden­ti­fies: (1) poorly-​prepared stu­dents, and (2) an over-​emphasis on enter­tain­ment and teach­ing rat­ings that moti­vates instruc­tors to offer sim­plis­tic lessons at the expense of sub­stan­tive learn­ing. The first is related to the pathetic under­grad­u­ate edu­ca­tions most folks receive and the sec­ond is, well, an exam­ple of an incen­tive prob­lem. (We’ll get back to both of these below.)

Incen­tive Prob­lems Are Easy to Iden­tify, but Dif­fi­cult to Solve

Incen­tives prob­lems are as nat­ural and as old as recorded his­tory: every­body wants what they want. In the Old Tes­ta­ment, were Adam and Eve any­thing if not incen­tive prob­lems? Cain? We could go, but there’s no rea­son. All of the indi­vid­u­als were free to act in a decen­tral­ized set­ting, and failed to live up to their responsibilities.

In the New Tes­ta­ment, Jesus dis­cusses incen­tive prob­lems on any num­ber of occa­sions. Two of our favorites: (1) the para­ble of the faith­ful and unfaith­ful ser­vants (Luke 12:41 — 48) and (2) the para­ble of the good shep­herd, (John 10:11 — 13). All con­sider the fallen nature of man and his (com­pletely nat­ural) self­ish behavior.

That being said, there is not a more com­plex topic to address in busi­ness schools – or any type of school, for that mat­ter – than incen­tives. That’s because the topic involves social (or multi-​party) sit­u­a­tions where one needs to be able to pre­dict how another party will respond autonomously and freely to con­trol mech­a­nisms like com­pen­sa­tion schemes.

Many of our read­ers already know that deci­sions can be cat­e­go­rized as games against nature – single-​person decision-​theory – and games against oth­ers, i.e., game the­ory. Gen­er­ally – though not pre­cisely – one can think of the inves­ti­ga­tions in the nat­ural sci­ences as exam­ples of single-​person deci­sions and inves­ti­ga­tions in the social sci­ences as exam­ples of multi-​person deci­sions, e.g., how does one respond to a sur­vey so how should the researcher inter­pret that response?

Incen­tive or agency prob­lems – and infor­ma­tion eco­nom­ics prob­lems in gen­eral – can often be mod­eled math­e­mat­i­cally using game the­ory or sim­i­lar meth­ods. In many of these prob­lems of inter­est to busi­ness stu­dents, one decision-​maker – say, the supe­rior or prin­ci­pal – is attempt­ing to max­i­mize his own expected sat­is­fac­tion or prof­its while ensur­ing that (1) the other per­son – the sub­or­di­nate or agent – is will­ing to par­tic­i­pate with him (in the social set­ting like a firm or orga­ni­za­tion) and (2) with full knowl­edge that the sub­or­di­nate or agent will do what’s best for himself.

Those two con­di­tions – par­tic­i­pa­tion and incentive-​compatibility – con­strain the principal’s abil­ity to max­i­mize his own expected sat­is­fac­tion, and the lat­ter prob­lem is espe­cially vex­ing to solve because it means that one of principal’s con­straints is the other person’s opti­miza­tion prob­lem. How do you do what’s best for your­self while real­iz­ing that the other per­son is also behav­ing oppor­tunis­ti­cally (by doing what’s best for himself)?

Objec­tively mod­el­ing these issues as math­e­mat­i­cal prob­lems tends to require a rather high level of sophis­ti­ca­tion, and solv­ing the resul­tant prob­lem – or even know­ing when a math­e­mat­i­cal solu­tion exists – requires an even greater under­stand­ing of advanced cal­cu­lus, opti­miza­tion, real analy­sis, and other math­et­i­cal the­o­ries and tech­niques.³

Very few MBA stu­dents are pre­pared to tackle those top­ics (and their appli­ca­tions) at that level of understanding.

Our Root Causes, Again

A larger set of stu­dents can han­dle sim­pli­fied illus­tra­tions and exam­ples of prob­lems that tend to be more numer­i­cal in nature. Often, when taught in con­junc­tion with a math soft­ware pro­gram, they can gain a keen under­stand­ing of the sub­tle issues that arise in the study of incen­tives, e.g., pay­ing more for more out­put isn’t nec­es­sar­ily opti­mal nor incentive-​compatible.⁴

Unfor­tu­nately, the root causes that we iden­ti­fied above – igno­rance and selfishness/​greed – make it dif­fi­cult for most instruc­tors to offer and suc­cess­fully teach such a course to MBA students.

We’ll empha­size the stu­dents’ igno­rance and not the instruc­tors’; instead, we’ll focus on their selfishness.

Most MBA stu­dents are poorly pre­pared to think clearly, abstractly, and quan­ti­ta­tively, and that makes it a chal­lenge to teach them either (1) quan­ti­ta­tive sub­jects or (2) top­ics that can be effec­tively mod­eled, illus­trated, or explained in a quan­ti­ta­tive manner.

Incen­tive prob­lems fall into the lat­ter cat­e­gory. (What we’d call) sim­ple math­e­mat­i­cal or numer­i­cal mod­els pro­vide (by def­i­n­i­tion) abstract illus­tra­tions of par­tic­u­lar phe­nom­ena and behav­iors. They’re rarely solu­tions to real world problems.

Most MBA stu­dents are not sophis­ti­cated enough to han­dle that dis­tinc­tion; they want recipes, not thought processes, and recipes are eas­ier to teach and grade. It’s not because the stu­dents are stu­pid, but it often is because they were poorly-​trained as under­grad­u­ates and in require, core classes. Per Mr. Jacobs’s essay, there’s gen­er­ally not much evi­dence of profs teach­ing compensation-​related recipes in busi­ness schools because of the lack of rel­e­vant incentive-​related courses. Thatt’s evi­dence of absence (of the courses), rather than an absence of evidence.

There’s much more evi­dence of that behav­ior in finance classes, where stu­dents want recipes for val­u­a­tion. They’ll take abstract mod­els, with either unre­al­is­tic assump­tions or very, very spe­cial­ized assump­tions and unwit­tingly (and unknow­ingly) treat them as very prac­ti­cal and pre­cise meth­ods that cal­cu­late the one true value of the thing.

Unfor­tu­nately, they’re often encour­aged to do so by their pro­fes­sors because it’s much eas­ier to teach numer­i­cal – though irrel­e­vant or mis-​specified – recipes than it is to teach (and grade) thought processes.

In fact, that ten­dency to dumb-​down teach­ing even extends to some fac­ulty mem­bers’ research agen­das. Dur­ing our aca­d­e­mic career, we attended any num­ber of sem­i­nars where we heard the pre­sen­ter jus­tify his or her overly-​simplistic and vac­u­ous model by argu­ing that “we want to be able to explain it to MBA students.”

Imag­ine if med­ical research were con­ducted in the same man­ner? Or any seri­ous field of inquiry for that matter?

From our per­spec­tive, it’s com­pletely ass-​backwards (and, in fact, its pres­ence goes par­tially to explain why we’re in the pri­vate sec­tor, today).

In an ideal words, the ped­a­gog­i­cal empha­sis would be on edu­cat­ing the stu­dents by attempt­ing to pull-​them-​up to a level that they had not antic­i­pated nor even known existed, and not pre­sent­ing dumb-​downed “research” papers for enter­tain­ment or pre­tense, but, hey, the lat­ter alter­na­tive is easy, and one can gen­er­ally gar­ner higher teach­ing rat­ings by not chal­leng­ing the stu­dents, espe­cially if that per­spec­tive and tech­nique is per­va­sive within the school. (We knew any num­ber of fac­ulty mem­bers at very expen­sive and seem­ingly pres­ti­gious insti­tu­tions who would pro­vide “sam­ple” or “prac­tice” exams before test dates – the actual exams would have slightly-​changed num­bers; who would sched­ule fre­quent guest speak­ers because “the stu­dents like it (and we don’t have to pre­pare);” and would show videos of fac­to­ries or what­ever once per week because, again, “the stu­dents like it (and we don’t have to pre­pare).” (Geez, it’s almost enough to make one cynical.)

Any­way, that com­bi­na­tion of poor prepa­ra­tion of most stu­dents and the mis­aligned incen­tives of b-​school pro­fes­sors make true learn­ing about these thorny and dif­fi­cult (social) prob­lems, which all firms and orga­ni­za­tions face, nearly impos­si­ble to achieve.

Why It’s Dif­fi­cult to Teach about Incen­tives Issues

It’s not just the math­e­mat­i­cal nature of the most com­pelling mod­els of incen­tives that makes teach­ing dif­fi­cult. It’s also because the prob­lems are not par­tic­u­larly robust. By that we mean, illus­tra­tions and exam­ples must be care­fully (and empa­thet­i­cally) con­structed, or they’re either (1) extremely stu­pid and un-​insightful, or (2) extremely spe­cial­ized, detailed, and so qual­i­fied (by assump­tions) that they need a very high degree of math­e­mat­i­cal under­stand­ing to com­pre­hend and solve (and they end-​up say­ing very lit­tle, anyway).

The fer­tile mid­dle ground requires instruc­tors and stu­dents to pos­sess a rather high level of eco­nomic rea­son­ing and strong math skills. We’ll avoid crit­i­ciz­ing instruc­tors, here, but unfor­tu­nately, many MBA pro­grams have de-​emphasized, elim­i­nated, or con­sol­i­dated micro­eco­nom­ics courses, and those courses are (or were) the best place to develop the req­ui­site level of eco­nomic rea­son­ing. In those courses and well-​designed incen­tives courses, there is no sub­sti­tute for a lot of hard work.

By the way, we unsuc­cess­fully tried to estab­lish just such a Con­trol & Incen­tives course at our last aca­d­e­mic employer, but there were no required econ courses and only a few very moti­vated, very curi­ous, or previously-​trained stu­dents would enroll in the elec­tive. (Too much work!) As a pub­lic ser­vice, we’ll attempt to put that course mate­r­ial on-​line in the near future.

But Dif­fi­culty Is Really No Excuse

It’s up to trustees and deans to ensure that schools and pro­fes­sors edu­cate MBAs, rather than attempt to be “pop­u­lar.” That’s true at both the indi­vid­ual level and the sum of the indi­vid­ual lev­els, i.e., the school level, where administration’s allow them­selves to be sub­jected to the whims of Busi­ness Week writ­ers and sur­vey respon­dents. As a fac­ulty mem­ber, we won our share of teach­ing awards while try­ing to do the right thing; so, there’s no sour grapes here, and we know that it can be done; how­ever, we sus­pect that the short-​term empha­sis will not change. There’s too much iner­tia and very lit­tle confidence.

From our self­ish per­spec­tive, it’s not as bad as it seems because that gen­eral fail­ure to learn and teach presents many oppor­tu­ni­ties for con­sul­tants who under­stand both incen­tives and risk – peo­ple like our­selves. (We’ve writ­ten exten­sively about both issues, espe­cially as they per­tain to the cur­rent finan­cial cri­sis. Please search the archives if you’re inter­ested. Our Illus­tra­tions dis­cuss many of these issues, too.)

Are you sure that your firm or orga­ni­za­tion isn’t about to do some­thing stu­pid with incen­tive pay or claw­backs or whatever?

We’ll likely con­tinue to revise and edit this post in the near future. (It’s long and there’s prob­a­bly a few typos, but then TQM is rarely optimal.)

Copy­right © 2009 Spero Consulting.

---

Foot­notes:

There are two related essays in the edi­to­r­ial sec­tion of today’s (March 30) edi­tion of The Wall Street Jour­nal regard­ing gov­ern­ment over­sight and reg­u­la­tion that are worth men­tion­ing: Wel­come, Busi­ness­men, to Gov­ern­ment Over­sight and Trans­parency Is More Pow­er­ful Than Reg­u­la­tion. We’ll men­tion the suf­fo­cat­ing nature of reg­u­la­tions and then dis­cuss the more inter­est­ing topic last, includ­ing our own work of using XML-​based sys­tems and tags for (inter­nal) man­age­ment infor­ma­tion sys­tems, which relates to the dis­cus­sion of XBRL sys­tems in the sec­ond column.

Suf­fo­cat­ing Reg­u­la­tions and Bureaucracy

Is there any other kind? Well, yes. As we see it, reg­u­la­tion is either suf­fo­cat­ing or inef­fec­tive, and the for­mer often has a crush­ing feel about it. Like mod­ern dig­i­tal tele­vi­sion sets, gov­ern­ment reg­u­la­tors seem to have no “fine-​tuning” dial; it’s gen­er­ally one extreme or the other: either there’s “NO EXCEPTIONS” or “it’s all good, do what you want.”

Vic­to­ria Toens­ing dis­cusses that over­bear­ing weight of the gov­ern­ment in Wel­come, Busi­ness­men, to Gov­ern­ment Over­sight in which she high­lights, among other indig­ni­ties, the silli­ness of gov­ern­ment offices unable to accept small gifts like cherry pies. Our guess is that she has spent most of her life in pub­lic ser­vice and doesn’t appre­ci­ate how sim­i­larly bureau­cratic large cor­po­ra­tions can be, but that’s besides the point because much – although not all – of that cor­po­rate bureau­cracy is induced by gov­ern­ment regulation.

We’ve dis­cussed both neg­a­tive aspects of reg­u­la­tion in numer­ous posts although we tend to high­light inef­fec­tive­ness because it has been very obvi­ous in the cur­rent finan­cial cri­sis and the mort­gage débâ­cle that pre­ceded (and which con­tin­ues to coin­cide with) it. For exam­ple, on Sat­ur­day we wrote The Cure is Worse than the Dis­ease, which crit­i­cizes Mr. Geithner’s pro­posed finan­cial sys­tem reg­u­la­tions.¹ We fear the suf­fo­ca­tion to come.

Gen­er­ally, we favor decen­tral­ized gov­ern­ment and much pre­fer decen­tral­ized work­ing envi­ron­ments, i.e, light reg­u­la­tion with the polic­ing authority’s option to crush, i.e., heav­ily penal­ize for indis­cre­tions. We take that from the Bible and the Para­ble of the Good (and Bad) Ser­vants, which cov­ers both moral haz­ard and igno­rance, and that’s why we’re strong pro­po­nents of nation­al­iz­ing the weak­est of the large banks. (Not because we think the gov­ern­ment will man­age them bet­ter but because we think share­hold­ers and cur­rent man­age­ments have for­saken the right to con­trol those assets.)

Suf­fo­cat­ing reg­u­la­tions and bureau­cracy usu­ally pro­vide no ben­e­fit to soci­ety and are inhu­mane and demean­ing. If “effec­tive,” they usu­ally end up killing the thing they are try­ing to pro­tect. (Nation­al­ized health-​care any­one?) In short, that’s why we’re against Mr. Geithner’s plan.

Trans­parency Anyone?

The other col­umn worth men­tion­ing is L. Gor­don Crovitz’s Trans­parency Is More Pow­er­ful Than Reg­u­la­tion in which he focuses his atten­tion on a sub­sti­tute for exten­sive reg­u­la­tory over­sight: more report­ing trans­parency. For sup­port of his posi­tion, he men­tions for­mer Supreme Court Jus­tice Louis Brandeis’s point that “sun­light is the best dis­in­fec­tant,” which we often cite, but is irrel­e­vant here.

While we tend to agree with many of his points, we think, that in the end, Mr. Crovitz draws the wrong con­clu­sions because (1) in gen­eral, in social set­tings more trans­parency isn’t nec­es­sar­ily bet­ter (doesn’t nec­es­sar­ily improve social wel­fare and can decrease it), and (2) in the spe­cial case of secu­ri­ti­za­tions of pooled assets, addi­tional trans­parency won’t solve the prob­lem of flawed pric­ing mod­els because the mod­els’ own­ers have lost con­fi­dence in them.

Is More Infor­ma­tion Always Better?

It depends. In sin­gle per­son games – i.e., nat­ural sci­ence exper­i­ments and games against nature, more infor­ma­tion is bet­ter. Roughly, that means it leads to higher expected sat­is­fac­tion for the par­tic­i­pant.² Clearly, record-​keeping is nec­es­sary for sev­eral rea­sons, but often those records don’t nec­es­sar­ily pro­vide mar­ginal ben­e­fit for decision-​makers in all deci­sions, i.e., the records might not iden­tify addi­tional rel­e­vant or dif­fer­en­tial costs or ben­e­fits among the pos­si­ble alter­na­tives for the deci­sion. Alter­na­tively, because they are not per­fectly ratio­nal, decision-​makers may not be able to cat­e­go­rize and syn­the­size or relate the new infor­ma­tion that is present, or they might mis­use it.

In a seem­ing con­tra­dic­tion to that view, last week, in Sep­a­rat­ing the Mort­gage Débâ­cle from the Liq­uid­ity Cri­sis, we agreed with Her­nando de Soto’s rec­om­men­da­tion that more details about con­tin­gent claims and secu­ri­ti­za­tion con­tracts should be made pub­lic, and Mr. Crovitz explains how this is tech­ni­cally fea­si­ble through the XBRL initiative.

As we see it, such details are infor­ma­tive about cer­tain aspects of the con­tracts, but not what Mr. Crovitz thinks. For exam­ple, it might help cred­i­tors bet­ter under­stand par­tic­u­larly low out­comes asso­ci­ated with cer­tain secu­ri­ties; so, we think that the details are worth report­ing, BUT the addi­tional details may not help with pric­ing claims on the pooled assets. Thus, we don’t see how trans­parency will induce liq­uid­ity. In fact, mar­kets often fail because there is “too much” trans­parency to sus­tain trans­ac­tions, i.e., no one wants the clearly-​identifiable crap – the lemons.

As we’ve writ­ten in the past, one of the prob­lems with these pric­ing mod­els for pooled assets is that their own­ers have lost con­fi­dence in them. They’ve lost con­fi­dence because they view the mod­els as no longer applic­a­ble, and they view them as no longer applic­a­ble because they have failed empirically.

They failed because they did not cap­ture the rela­tion­ships and inter-​relations among the assets, par­tic­u­larly among res­i­den­tial mort­gages. (In other words, the traders and ana­lysts vastly under-​estimated the joint depen­den­cies among cash flows and col­lat­eral val­ues, which those folks may express as hav­ing a poor esti­mate of the cor­re­la­tions, but which is likely more com­pli­cated and far less cal­cu­la­ble than that.) We’ve writ­ten about that on sev­eral occa­sions, includ­ing here: Trad­ing, Incen­tives, Orga­ni­za­tional Struc­ture and Risk Man­age­ment, where we explain it as a con­ta­gion. (We also dis­cuss it in Well, This Is a Fine Mess You’ve Got­ten Us into…. along with other still per­ti­nent issues.)

The prob­lem is that there are few math­e­mat­i­cally tractable ways to spec­ify how these assets are related; so, solv­able – but non­de­scrip­tive and mis­spec­i­fied – meth­ods were employed. In sta­ble times and with a bit of good luck, that mis­spec­i­fi­ca­tion didn’t seem to mat­ter. Unfor­tu­nately, luck changed, and did and it does now.

So, we don’t see how trans­parency will induce trad­ing, but that doesn’t mean that trad­ing can­not occur. (Mr. Crovitz has a good solu­tion, but to a dif­fer­ent prob­lem, i.e., Mr. de Soto’s problem.)

Our Solu­tion

Since Sep­tem­ber we’ve rec­om­mended changes in tax poli­cies – via mort­gage invest­ment tax cred­its or imme­di­ate write-​offs of pur­chase prices – as a way to induce trade and cre­ate liq­uid­ity in these secu­rity mar­kets. Pro­vid­ing a 30 – 40% cush­ion in the pur­chase price, will induce trad­ing even if buy­ers aren’t com­pletely con­fi­dent of their cal­cu­la­tions. Imag­ine if the same tax incen­tives were avail­able to new car buy­ers? (See “The Good Cop” sec­tion of Poor Mr. Gei­th­ner: No For­est, No Trees, Just Lost for a recent overview of our plan.)

What Does This Have to Do with MIS?

The same types of sys­tem that XBRL is based upon are avail­able very cheaply for inter­nal decision-​makers. We’re design­ing and imple­ment­ing sim­i­lar robust, tagged sys­tems for our clients. They are eas­ily search­able sys­tems – both infor­mally (ad hoc) and for­mally (rou­tine reports); they’re easy to update and edit; they’re secure; and they’re rel­a­tively inex­pen­sive. The ben­e­fits of tech­nol­ogy can now be real­ized by any size firm or orga­ni­za­tion. Con­tact us for more information.

As always, we might update this post after we re-​read it.

Copy­right © 2009 Spero Consulting.

---

Foot­notes:

No, dear reader, this is not about the usual sus­pects: the banks, the Pres­i­dent nor Con­gress. It’s about a silly essay in the Taste sec­tion in today’s edi­tion of The Wall Street Jour­nal. In par­tic­u­lar, we mean the one enti­tled, The Real March Mad­ness by Richard Ved­der and Matthew Denhart.

As we read it, Ved­der and Den­hart com­plain that col­le­giate ath­let­ics are not cod­dled or treated well enough and that they should con­sider union­iz­ing and bar­gain­ing for greater ben­e­fits. You know, ben­e­fits other than the free, if gen­er­ally under-​utilized tuition and edu­ca­tion; free room-​and-​board; free travel; free use of state-​of-​the-​art facil­i­ties; free health-​care; and the adulation.

Now before con­tin­u­ing, it’s worth not­ing that we’re rarely con­fused with apol­o­gists for the NCAA or fanat­i­cal col­lege pres­i­dents or alumni. Our PhD is from a small, rig­or­ous uni­ver­sity, and our first aca­d­e­mic posi­tion was at a sim­i­lar school. Sub­se­quently, we spent a few years at a large, state insti­tu­tion that over-​emphasized sports. (No, it wasn’t prison, but there was no short­age of back-​stabbers, shanks or prison bee-​atches.)

While at the state school, we saw what we had missed in our prior Divi­sion III envi­rons: the over-​emphasis on sports. (Now, our Alma mater seems to be overly-​fixated on its incred­i­bly shrink­ing endow­ment. Per­haps if they had focused more on sports, they would have done less dam­age to them­selves.) While we write that par­en­thet­i­cally and as a joke, we’re quite seri­ous about the converse.

At the state school, we real­ized that if in the administration’s judge­ment, the most impor­tant areas to focus their atten­tions were sports and sport facil­i­ties, which really are inessen­tial to the true pur­pose of higher edu­ca­tion, then work­ing on those tasks were, in fact, the very best places to focus their atten­tion. In other words, given their poor judg­ments would you really want them to focus on edu­cat­ing the youth? Or in still other words, thank God for decen­tral­ized insti­tu­tions that basi­cally run themselves.

There are any num­ber of crit­i­cisms to levy against Ved­der (and his side­kick), and we’ll men­tion as many as we can – until the Pitt game starts.

First, Mr. Ved­der should put his money where is mouth (or key­board) is. Per his allu­sion to minor league base­ball, he should cre­ate minor league bas­ket­ball and foot­ball leagues and pay the young ath­letes at least the “liv­ing” wage that he proposes.

How­ever, if that is infea­si­ble, then focus on imple­mentable, real­is­tic recommendations.

Secondly, he writes as if the ath­letes have no choices in the mat­ter or respon­si­bil­i­ties for their own actions and deci­sions. Yes, we’re sure that play­ing major col­lege ath­let­ics requires a HUGE time com­mit­ment, BUT it’s fun, AND in almost every case, the play­ers like to play sports; so, they do get sat­is­fac­tion from the experience.

More­over, mil­lions and mil­lions of stu­dents have attended col­lege and grad­u­ate school while work­ing full-​time. Many of those indi­vid­u­als have enrolled in school full-​time while also work­ing full-​time. While not every­one is suc­cess­ful, some­how many of those indi­vid­u­als found a way to com­plete their stud­ies with­out the assis­tance from team tutors and counselors.

In the end, despite the time com­mit­ments and with the assis­tance, the student-​athletes are still respon­si­ble for them­selves, and IF their aca­d­e­mic per­for­mance is worst than other stu­dents, it’s their per­for­mance and their respon­si­bil­ity. There is freewill.

Thirdly, it seems that Mr Ved­der is apply­ing the labor the­ory of value; the ath­letes – because they play the games – rather than the uni­ver­si­ties which orga­nize them, should earn the rents. Here, as counter-​intuitive as it seems, the uni­ver­si­ties are play­ing the roles of the entre­pre­neurs, and if they can meet the ath­letes’ reser­va­tion util­ity – i.e., par­tic­i­pa­tion con­straint – and incen­tive com­pat­i­bil­ity con­di­tion, and they do, then why should they over­pay? (We mean that there are no short­age of will­ing ath­letes and they cer­tainly seem to try their hard­est and best, so why should the schools do any­thing dif­fer­ently, eh Mr. Marx, err, Vedder?)

Finally – it’s almost tip-​off – the authors make a fuss about the fact that recent grad­u­ates (and non-​graduates) make sub­stan­tially more play­ing pro sports than they do in col­lege. That’s irrel­e­vant, but it is also true for a host of occu­pa­tions, whether the train­ing involves the class­room work or appren­tice­ships in the trades.

As an exam­ple, they use Kevin Durant and his pro salary com­pared to his non-​existent one in his sin­gle year of col­lege. It seems highly likely that if he could have played in the pros at the same salary with­out going to col­lege he would have. More­over, it’s highly likely that he con­sid­ers his unpaid year at Texas as well worth the time and effort given the oppor­tu­nity that he had to exhibit his tal­ent and skills and his sub­se­quent salary.

As we said, abject silli­ness. Shame on The Wall Street Jour­nal for pub­lish­ing it.

We’ll likely update and edit this post after the game.

Or Cen­tral Plan­ning as a Mar­ket Solution

We saw in yesterday’s (Feb­ru­ary 4^th) edi­tion of The Wall Street Jour­nal that cer­tain leg­is­la­tors, includ­ing Bar­ney Frank, want a gov­ern­ment agency, pos­si­bly the Fed­eral Reserve, to “con­trol” sys­temic risk in the econ­omy, par­tic­u­larly in the finan­cial markets.

We’ll ignore the fact that this is the same Bar­ney Frank who induced much sys­temic risk by insist­ing for many years that Fan­nie Mae and Fred­die Mac make home own­er­ship afford­able for those who could not afford a home. He was then shocked, shocked, and dis­mayed that a good per­cent­age of those folks couldn’t afford their new homes. Yes, very sur­pris­ing, indeed!

Doing more harm than good: Instead, we’re writ­ing because we find it quite ironic that an agency, i.e., a sin­gle gov­ern­ment reg­u­la­tor or a small group of reg­u­la­tors would be able to “con­trol” and “man­age” some­thing like sys­temic risk with­out either (1) com­pletely destroy­ing the econ­omy they’re assigned to pro­tect or (2) con­vert­ing their own idio­syn­cratic per­spec­tives and pref­er­ences into more or new kinds of sys­temic risk.

Good inten­tions and the road to hell: The first out­come is actu­ally the worst-​case sce­nario of the sec­ond one and isn’t much dif­fer­ent than Mr. Frank con­vert­ing his own idio­syn­cratic pref­er­ences about home own­er­ship into the gigan­tic mort­gage losses incurred by Fan­nie and Fred­die, among oth­ers. One need not be greedy or self­ish to be misguided.

By far, the eas­i­est way – and the historically-​proven way – to con­trol sys­temic risk would be to destroy the econ­omy. That would cer­tainly elim­i­nate vari­a­tions – the ups and down – because the ups would be gone: kind of like the for­mer Soviet Union or modern-​day Cuba.

We’re sure that the destruc­tion would be inad­ver­tent and would be the out­come of well-​intentioned efforts, but that wouldn’t lessen the pain.

We ask: which past (and failed) attempt at cen­tral plan­ning has not been about “pre­serv­ing jobs” or “cre­at­ing jobs” or doing some­thing won­der­ful for human­ity? We can’t think of any.

The irony of sys­tem­atiz­ing idio­syn­cratic risk: As we men­tioned above, our point is that cen­tral­iz­ing decision-​making in one per­son or small group of peo­ple and per­mit­ting them to reg­u­late or gov­ern the econ­omy cre­ates addi­tional sys­temic risk to the detri­ment of all.

We have dis­cussed these issues in a num­ber of posts, includ­ing Com­mon Sense? Smart Money? Oh, Please! 

We’ve focused on the notion that the idio­syn­cratic becomes the sys­temic as port­fo­lios get larger and the decision-​making becomes more cen­tral­ized, and we’ve men­tioned it quite often because it is gen­er­ally ignored by folks. Such risk is assumed-​away in intro­duc­tory finance mod­els that show ben­e­fits of diver­si­fi­ca­tion; so, most folks don’t think about it.

We men­tioned it when dis­cussing merg­ers in Big­ger Is Not Nec­es­sar­ily Bet­ter:

“Each senior decision-maker’s idio­syn­cratic (and pos­si­bly irra­tional) beliefs and judg­ments affect a larger and larger share of the economy’s resource deci­sions, and that can’t be a good thing. Thus, there is a trade-​off of the cost sav­ings (of con­sol­i­da­tion) ver­sus the addi­tional risk of such centralized decisions.”

Think of it as the undi­ver­si­fi­able risk due to the fact that the port­fo­lio is cho­sen by a semi-​rational human or small group of humans, each with their own unique and shared flaws and assump­tions. The fact that such an error term does not exist in these finan­cial mod­els does not mean it is absent. It means that the model is an abstract, stream-​lined ver­sion of real­ity that ignores cer­tain fac­tors – often­times, impor­tant factors.

As we’ve often men­tioned, given our con­ser­v­a­tive nature, we do wish our elected lead­ers and appointed reg­u­la­tors would take an equiv­a­lent of the Hip­po­cratic Oath: beyond all else, “do no harm.” Unfor­tu­nately, as their actions over the past sev­eral months have shown, that is ask­ing far too much of them.

So we ask: can’t we all just wear our “WIN” but­tons from the sev­en­ties? We can change the “I” from “infla­tion” to “illiq­uid­ity” to “Whip Illiq­uid­ity Now!” It would be silly today as it was when Ger­ald Ford was Pres­i­dent, but it would be far less harm­ful than hav­ing a cou­ple geniuses – like, say, Bar­ney Frank or Henry Paul­son – sort through and “solve” our problems.

The Wall Street Jour­nal has an edi­to­r­ial in today’s paper – Jan­u­ary 14 – that seems to be ripped from our head­lines: it calls for the dis­mem­ber­ment of Cit­i­group, and it implies that Citi has lost its right to exist. (See When Is Enough Enough?, for exam­ple, or any of our calls to nation­al­ize it.)

As we’ve seen in var­i­ous news reports, Cit­i­group has lost about $30,000,000,000 or so in the last five quar­ters and has received about $45,000,000,000 in TARP funds, and the fed­eral gov­ern­ment has guar­an­teed another $250,000,000,000 or so of its debts.

And yet, and yet, Citigroup’s stock price is about $5, which gives it a mar­ket value, accord­ing to Google Finance of about $32 bil­lion. That’s less than 10% of its share price two years ago and about 20% of its share price this time last year.

As a point of com­par­i­son, if the fed­eral gov­ern­ment gave us $45 bil­lion, we would be worth $45 bil­lion. (Well, almost $45 bil­lion, but a lot closer to $45 bil­lion than $32 bil­lion. And, yes, we know there is a dif­fer­ence between the government’s pre­ferred invest­ment and mar­ket value of the com­mon shares.)

Hmmm, with­out both­er­ing to check the tax impli­ca­tions, let’s gross-​up the loss of about $30,000,000,000 to the $45 bil­lion. That means that the gov­ern­ment has sub­si­dized all of the rec­og­nized losses to date.

So, despite the guar­an­tee of debt, which could be val­ued the same way that banks esti­mate val­ues of their insured deposits, and despite the addi­tional deposit insur­ance cov­er­age, etc., soci­ety and the world econ­omy think that Cit­i­group isn’t worth a whole lot.¹

Dili­gent, and younger read­ers with good mem­o­ries, may recall that as far back as Sep­tem­ber we sep­a­rated the mort­gage fiasco from the larger, and far more seri­ous, liq­uid­ity cri­sis in con­fi­dence. (Here’s an entry from early Octo­ber: Even A Per­fect Bailout Will Fail.)

We cite Cit­i­group as prima facie evi­dence of that dis­tinc­tion. Based upon equity val­ues – despite the government’s mas­sive injec­tion of funds and its guar­an­tees – we’d say that the mort­gage fiasco has informed investors through­out this coun­try and across the world that’s Citi’s man­age­ment excels at value destruc­tion, and that’s the con­sen­sus prospec­tive esti­ma­tion. That is, of course, unless investors esti­mate that rec­og­nized losses, which appear on finan­cial state­ments, are only a frac­tion of Citigroup’s true losses so far.

This wouldn’t be the first time that Cit­i­group under-​estimated its losses. As the Jour­nal edi­to­r­ial notes, in Octo­ber, 2007, Citi offi­cials claimed that it had only “$70 mil­lion in indi­rect expo­sure to sub­prime assets.” Now, how many orders of mag­ni­tude is that from the truth? So whether clue­less or duplic­i­tous, “why trust them?” the mar­ket seems to be saying.

In this case, it seems hard to argue with that logic.

By the way, the front page head­line of today’s paper is “Cit­i­group Ready to Shrink Itself by a Third.” We won­dered – in jest – why the sec­ond line didn’t read, “In Small Attempt to Align Assets with Equity Values.”

Like always, we may edit this post in the future, in case our early-​morning, frost­bit­ten fin­gers have erred.

Copy­right © 2009 Spero Consulting.

---

Foot­note:

The Wall Street Jour­nal has an arti­cle today enti­tled, Mack and Thain Lose ’08 Bonuses.

We’re nei­ther sym­pa­thetic nor antag­o­nis­tic towards Mr. Thain, who has only been in his posi­tion for a year; so, we take no glee in his being shut-​out. Hope­fully, he’ll be able to make-​do with his $750,000 salary, $15-$20 mil­lion sign­ing bonus from late 2007, and his other accu­mu­lated wealth from his past exec­u­tive positions.

What inter­ests us in the arti­cle is the men­tion that Mor­gan Stan­ley plans to imple­ment com­pen­sa­tion schemes that include “claw back” fea­tures. That means that in the future, the firm could recoup ear­lier bonuses if, say, a trader later blows up.

Please note that we are writ­ing in gen­er­al­i­ties and not attempt­ing to con­struct an opti­mal con­tract, but we do see claw-​back fea­tures as mov­ing in the right direc­tion for both firms and employ­ees. (We’ve writ­ten pos­i­tively about sim­i­lar fea­tures before.)

At first glance, such claw­backs may seem to impose more risk on employ­ees, but if they’re struc­tured and used prop­erly, they need not; thus, we’d expect them to be wealth-​maximizing for the firm and expected-​utility max­i­miz­ing for employ­ees. That’s if they are con­structed intel­li­gently.

We’d hope that Morgan’s scheme is so con­structed – to, say, claw back por­tions of a 2008 bonus because trades or invest­ments made in 2008 sub­se­quently go bad.

We hope that the firm does not attempt to claw back a por­tion of say, a 2008 bonus because the trader made a money-​losing trade in 2009. We under­stand the aver­ag­ing effects of long-​term con­tracts, but believe that such rep­ri­mands would likely be per­ceived as being arbi­trary and capri­cious and sub­jec­tive and would likely have two effects: (1) before-​hand, many traders would leave to join hedge funds or to trade for them­selves, and (2) those traders who did stay and win large bonus awards could be expected to become sub­stan­tially more risk-​averse in the future (because both the cur­rent period’s bonus and past bonuses were all still at stake). In gen­eral, it doesn’t seem that most trad­ing and invest­ing firms want to induce traders to min­i­mize risk; instead, it is to man­age risk intel­li­gently or effi­ciently. If the goal were, in fact, to min­i­mize risk, then pay­ing a bonus as a func­tion of prof­its would be a huge mis­take in the first place. There’s much cheaper ways to induce that behavior.

The Good: Besides claw backs, we’d rec­om­mend that firms con­tinue to pay bonuses on earn­ings even after traders have left the firm – solely to induce them to behave and act in the firm’s long-​term inter­ests while they are employed. It is very tempt­ing to want to pun­ish for­mer employ­ees for leav­ing or for a vari­ety of real or per­ceived trans­gres­sions, but it is not nec­es­sar­ily the wis­est pol­icy nor fidu­cia­r­ily responsible.

Unfor­tu­nately, it seems that UBS may have taken that course.

We’re very grate­ful that the WSJ arti­cle men­tions that UBS imple­mented claw­backs in mid-​November because we had pre­vi­ously missed that announce­ment in the press.

The Bad: In August, we com­mented on UBS’s plans to use phan­tom shares in its com­pen­sa­tion schemes in Incen­tives at UBS and in Gen­eral. That plan seemed to impose a sub­stan­tial – we mean exces­sive – amount of risk on its employ­ees. W would strongly encour­age inter­ested par­ties to read that post.

From our read­ing of a few arti­cles more recent arti­cles, espe­cially this Lon­don Times arti­cle, UBS’s plan seems down­right vin­dica­tive. While that may be jus­ti­fied in the cases of for­mer senior exec­u­tives and while it may be sat­is­fy­ing to stiff employ­ees in bad times, it’s gen­er­ally not wealth-maximizing; it seems quite sub-​optimal.

UBS calls a neg­a­tive bonus a “malus.” Get it? It sub­sti­tutes “mal” for “bon” to get the oppo­site. Very clever!

The Ugly: Accord­ing to a Tele­graph arti­cle, UBS will attempt to claw back pre­vi­ously awarded, but not dis­trib­uted bonuses, if the bank under-​performs, and it could recover up to two-​thirds of the cash por­tion, which would be held in escrow for at least a year. So imag­ine that you, Joe Trader, or more pre­cisely Josef Trader, had a par­tic­u­larly good year in 2009, but the firm had com­pletely hor­ri­ble year in 2010; so, not only do you not get a 2010 bonus, but your 2009 bonus is gone, gone, gone. How would you feel? What are the odds that it could occur? Is it worth tak­ing the chance (bear­ing the risk) of such per­sonal losses? If it’s not, you may want to seek employ­ment elsewhere.

The Times arti­cle men­tions that Share-​based bonuses won’t vest for three years and exec­u­tives will be required to retain 75% of those shares for sev­eral more years, and the “malus” sys­tem will apply to shares, too. As we wrote in August and repeated above, such plans impose sub­stan­tial risk on employ­ees. UBS should expect to pay higher com­pen­sa­tion on aver­age and expect an exo­dus of employ­ees. We’d guess that it would lose many of its best, most con­fi­dent employ­ees, and many of its most risk-​averse, and espe­cially the inter­sec­tion of the two. Would you, dear reader, tol­er­ate such a scheme?

By the way, for exit­ing employ­ees, all bonuses paid on depar­ture will be sub­ject to the “malus” sys­tem. What are the chances that will be manip­u­lated against the employee (as, say, a short-​term way to boost current-​period profits).

In that regard, we love this quote from the bank that appeared in the Tele­graph arti­cle: “This should pre­vent any pay­ments that prove to be inap­pro­pri­ate in the near future.” But, when did pre­vent­ing any, which we take to mean “all” inap­pro­pri­ate, pay­ments become the goal?

In eco­nomic mod­els, profit-​maximization in the short-​term or wealth-​maximization in the long-​term do not imply the all costs can be elim­i­nated. We, and every other econ­o­mist that we know, teach that there is an eco­nomic level of costs that max­i­mizes prof­its. Like­wise, in decen­tral­ized orga­ni­za­tions, all dys­func­tional behav­ior can­not be elim­i­nated with­out also elim­i­nat­ing the ben­e­fits of auton­omy; it is throw­ing the prover­bial baby out with the bath-​water or being penny-​wise and pound-​foolish. (See just about any­thing that we’ve writ­ten in our Illus­tra­tions and Fal­lac­ies sec­tions, espe­cially about extrem­ists in Com­mon Man­age­r­ial Mis­takes in Decen­tral­ized Orga­ni­za­tions.)

We what find to be espe­cially galling is the fact that intelligently-​applied claw­backs are a great idea for both firms and employ­ees, but unfor­tu­nately, if (as an early adopter) UBS botches its imple­men­ta­tion – which given the infor­ma­tion in the press seems highly likely – then other firms will likely be hes­i­tant to use them. That’s a shame.

If large firms want to elim­i­nate risk, then we encour­age to elim­i­nate pro­pri­etary trad­ing and oper­ate rel­a­tively low-​risk, low-​margin businesses. That’s what we’ve rec­om­mended for government-​insured firms in our aptly-​titled post Elim­i­nate Pro­pri­etary Trad­ing at Insured Insti­tu­tions.

We’ll likely edit and add to this post in the near future.

Copy­right ©2008, Spero Con­sult­ing Incorporated.

In the thirty-​five years since the first “energy cri­sis,” have Chrysler, Ford, or GM avoided a sin­gle man­age­ment fad?

Have their col­lec­tive man­age­ments through the years embraced of any sin­gle fad that led to sus­tain­able improve­ments anywhere?

Now, it is true that many fads – and we are using that word pejo­ra­tively – con­tain use­ful rec­om­men­da­tions and are con­sis­tent with effec­tive and effi­cient man­age­ment. How­ever, that’s only if such poli­cies and tech­niques are thought­fully applied to one’s par­tic­u­larly orga­ni­za­tion and sit­u­a­tion, and that is a big, bold IF. 

Spend­ing vast amounts of time and energy try­ing to get apply the inap­plic­a­ble might be hard and expen­sive work and might require cre­ativ­ity and inge­nu­ity, but it is almost always worth­less, regard­less of the sat­is­fac­tion felt by com­plet­ing a dif­fi­cult project.

It’s our view that the thought­ful appli­ca­tion of sound busi­ness poli­cies and prac­tices pre­cludes the neces­sity for such fads in the first place. A dif­fer­ent per­spec­tive and spe­cial­ized exper­tise are ben­e­fits that some con­sul­tants offer, but the whole­sale revamp­ing of only cer­tain func­tions is usu­ally myopic and often as sense­less as fit­ting square pegs in round holes. The fact that XYZ worked for firm RST in indus­try LMN means very lit­tle for firm ABC in indus­try DEF unless there are direct analogues.

Like­wise, that fact that RST’s mar­ket value increased when it imple­mented XYZ means very lit­tle for firm ABC. For exam­ple, one could ask: how were equity mar­kets and indus­try indices mov­ing in gen­eral dur­ing that the time of the imple­men­ta­tion? As we have all read many, many times, cor­re­la­tion is not cau­sa­tion, and cor­re­la­tion based upon a sam­ple of one gen­er­ally doesn’t mean much, either.

In every func­tional area – pro­duc­tion, finance, sales, human resources – from qual­ity cir­cles and total qual­ity man­age­ment to supply-​chain man­age­ment and just-​in-​time inven­to­ries to process engi­neer­ing – even robot­ics – to activity-​based cost­ing to EVA™ to knowl­edge man­age­ment et cetera, et cetera, did the car-​makers avoid or ignore a sin­gle one of them? (We’ve actu­ally for­got­ten many older fads or the list could have been longer. We think that qual­ity cir­cles were after disco, but our mem­ory fails.)

Were any fad­dish tech­niques thought­fully applied? Was the wheat, so-​to-​speak, ever sep­a­rated from the mar­ket­ing chaff? Did the suc­cess of the imple­men­ta­tion (of a pre­vi­ous fad) pre­clude the adop­tion of a new one? Were any of the pro­grams true man­age­ment innovations?

We won­der how much have the “Big Three” spent on large con­sult­ing firms that were mar­ket­ing such fads and imple­men­ta­tions dur­ing those thirty-​five years? To what ben­e­fit, and who mea­sured the ben­e­fit? The con­sult­ing company?

Was the down­fall of the big three inevitable? We don’t think so. 

Was the time to col­lapse length­ened or has­tened by the pur­chase of those fads? Could any­thing have been sub­sti­tuted for in place of those fads that could have pro­vided longer, more per­sis­tent ben­e­fits? Say, some­thing like thought­ful, dis­ci­plined man­age­ment with a solid under­stand­ing of the busi­ness of design­ing, man­u­fac­tur­ing, and sell­ing automobiles?

We doubt that one per­son could pos­sess all of the req­ui­site knowl­edge to mas­ter and directly con­trol all oper­a­tions and func­tions of a large auto man­u­fac­turer, but that’s not the point. If senior man­age­ment does not have the per­sonal knowl­edge to design, build, and sell cars, then it needs the man­age­r­ial knowl­edge and dis­ci­pline to orga­nize and con­trol activ­i­ties of those who do. That’s not to be out­sourced through a sequence of fads.

With­out man­age­r­ial knowl­edge and discipline, what hope – other than good for­tune – does a firm, its employ­ees, and its share­hold­ers have to (1) avoid waste­ful fads and other mis­steps in the short-​term or (2) survive in the long-​term? This cer­tainly seems to be true for firms with 60% mar­ket share in the late 1960s, and 50% mar­ket share in 1980, and 35% as recently as fif­teen years, ago, and about 20% today. (Of course, we’re speak­ing of GM, the big three still had nearly 60% of the domes­tic mar­ket as recently as 2003.)

What hope does a gov­ern­ment bailout offer for insti­tu­tions that have squan­dered so much?

Does any­one else imag­ine that Henry Ford is spin­ning in his grave?