Trending Toward Inanity

Mark Penn’s new book, Microtrends, is so epically awful that it could take the entire polling industry down with it

Ezra Klein

If you want­ed to ruin the polit­i­cal career of Mark Penn, Hillary Clin­ton’s chief poll­ster and strate­gist, here would be one way to do it: First, cre­ate some sort of arti­fact bear­ing his name that you could use to tank his rep­u­ta­tion. A book would do per­fect­ly. Title it some­thing buzz-wordy and super­fi­cial, like Microtrends, though per­haps that’s too heavy-hand­ed. Fill it with vapid koans, like small is the new big” and the biggest move­ments in Amer­i­ca today are small.” To make it seem authen­tic, you’d want to ape Pen­n’s long-stand­ing affec­tion for com­bin­ing demo­graph­ic sala­mi-slic­ing with cutesy-nam­ing (this is the man who foist­ed Soc­cer Moms” upon our weary lex­i­con), mak­ing each short chap­ter an expo­si­tion of ever-more absurd groups – think Archery Moms,” Old New Dads,” and Ardent Ama­zons.” Final­ly, assert their impor­tance through wild and empir­i­cal­ly unsup­port­ed spec­u­la­tions. That last would be the key: You’d want the method­ol­o­gy so wild and slip­shod, so trans­par­ent­ly flawed, that no one would trust the ana­lyst ever again.

Aston­ish­ing­ly, Penn him­self has done exact­ly this. His new book Microtrends is so bad that the ques­tion – in a fair world – isn’t whether it will destroy his own rep­u­ta­tion, but whether it is so epi­cal­ly awful as to take the entire polling indus­try down with it.

First, a bit of full dis­clo­sure: Unlike every­body else in Wash­ing­ton, I have nev­er met Mark Penn. This, I am assured, is to my endur­ing dis­cred­it, as he’s appar­ent­ly a love­ly indi­vid­ual, and if I only knew him, I would under­stand that his pro­tec­tion of a union-bust­ing divi­sion with­in Bur­son-Marsteller, the PR firm of which he is the CEO, isn’t evi­dence of anti-union feel­ings at all. Nor is Pen­n’s cease­less advo­ca­cy for a cau­tious, hawk­ish, pro-cor­po­rate, don’t-rock-the-boat Demo­c­ra­t­ic Par­ty a func­tion of his beliefs, cor­po­rate back­ground or clients. Instead, it’s mere­ly The Num­bers.” Indeed, noth­ing Penn says or does can be ques­tioned, because he’s just there to give us The Num­bers. His per­son­al thoughts are immaterial.

At the same time, his per­son­al thoughts mat­ter. As the Wash­ing­ton Post recent­ly report­ed: In the four months since Clin­ton offi­cial­ly became a can­di­date, Penn has con­sol­i­dat­ed his pow­er, accord­ing to advis­ers close to the cam­paign, tak­ing increas­ing con­trol of the oper­a­tion. Armed with volu­mi­nous data that he col­lects through his pri­vate polling firm, Penn has become involved in vir­tu­al­ly every move Clin­ton makes, with the result that the cam­paign reflects the chief strate­gist as much as the can­di­date.” Even there, though, the Penn mys­tique per­sists: Pen­n’s pow­er, we’re told, comes from his volu­mi­nous data,” not his opin­ions. To argue with Penn is to argue with The Num­bers. And you’re not against Num­bers, are you?

That’s the Penn defense, and he and his friends have long stuck to it. Mark is some­body who is very, very com­fort­able with quan­tifi­ca­tion,” enthused Doug Schoen, his polling part­ner of over 30 years. He is very com­fort­able with num­bers.” It is this rep­u­ta­tion that, so far as I can tell, Mark Penn has writ­ten Microtrends to dis­pel. Unlike most poll­sters, Penn nev­er releas­es his raw num­bers, only his analy­sis. So we must take it on faith that his method­ol­o­gy is rig­or­ous, his polls accu­rate and his inter­pre­ta­tions fair. This book is our first oppor­tu­ni­ty to observe, at length, how adroit­ly Penn han­dles raw data. And the answer is stun­ning, even to a doubter like me. Mark Penn can­not han­dle num­bers. If this book were turned in as the final to an entry-lev­el sta­tis­tics class, Penn would not only be failed, but the pro­fes­sor might well retire in shame.

I first flipped through Microtrends while at the Year­lyKos con­ven­tion, and Penn, aston­ish­ing­ly, seemed to com­pre­hend the impor­tance of the loose­ly con­nect­ed, grass­roots-dri­ven, pro­gres­sive move­men­t’s flow­er­ing. I sus­pect the lefty boom will bring a surge in the pro­mo­tion of sheer cre­ative ener­gy,” Penn writes, dri­ven by an idea that is at the heart of this book – that small groups of peo­ple, shar­ing com­mon expe­ri­ences, can increas­ing­ly be drawn togeth­er to ral­ly for their inter­ests.” I was shocked – Penn was speak­ing admirably of left­ies,” not try­ing to recast them as mod­er­ates, not try­ing to write them out of the par­ty? He was endors­ing open-source pol­i­tics, rather than a top-down struc­ture? I had mis­judged the man!

I read on. Penn was talk­ing about actu­al left­ies – peo­ple who are born left-hand­ed. Increas­ing­ly grim, I absorbed the first hard blows of Pen­n’s inter­pre­ta­tive tech­nique: More left­ies,” he enthus­es, could mean more mil­i­tary inno­va­tion: Famous mil­i­tary lead­ers from Charle­magne to Alexan­der the Great to Julius Cae­sar to Napoleon – as well as Col­in Pow­ell and Nor­man Schwarzkopf – were left-hand­ed.” He uses the same thun­der­ing­ly awful log­ic to argue that we’ll see more art and music greats, more famous crim­i­nals, more great come­di­ans, more exec­u­tive great­ness,” and bet­ter ten­nis and bas­ket­ball players.

This is what sta­tis­ti­cians – or any­one who has tak­en a sta­tis­tics class – call a correlation/​causation error.” It is not enough to cher­ryp­ick a cou­ple famed mil­i­tary lead­ers, notice that they’re left­ies and assume that some­thing intrin­sic to their hand­ed­ness caused their tac­ti­cal genius. It is not enough to say that past cul­tures dis­cour­aged left-hand­ed­ness and use that as a stand-in for dis­cour­ag­ing cre­ativ­i­ty of all sorts. To say that Bill Gates is right-hand­ed does not sug­gest that a greater pro­por­tion of right-hand­ed peo­ple would mean more Bill Gate­ses. For a pro­fes­sion­al poll­ster to imply that cor­re­la­tion equals cau­sa­tion is like a fire­fight­er try­ing to put out flames by toss­ing a toast­er into the blaze – it bespeaks a com­plete unfa­mil­iar­i­ty with the rel­e­vant techniques.

What’s more amaz­ing is this: A page ear­li­er, Penn argues that the rise in left­ies has noth­ing to do with there being more left­ies, and every­thing to do with more per­mis­sive par­ent­ing. In oth­er words, where chil­dren used to be trained out of left-hand­ed­ness, now par­ents shrug their shoul­ders, say­ing it’s okay.” So not only does Penn fail to prove that left­ies are genet­i­cal­ly dif­fer­ent in some impor­tant way, he also sug­gests that the gene pool is no dif­fer­ent, and that there are as many of them around now as always. It’s a fal­la­cy atop an error built around some­thing that isn’t hap­pen­ing.

This isn’t an iso­lat­ed exam­ple. In a chap­ter called Aspir­ing Snipers,” Penn explains, It’s the rare moment when a poll stops me in my tracks and reori­ents my under­stand­ing of things.” One such poll was con­duct­ed last fall, when Ben­dix­en and Asso­ciates asked 601 young Cal­i­for­ni­ans what they’d be doing in 10 years. About 1 per­cent – so, a hand­ful – said they’d be snipers. Cer­tain­ly, that’s an odd reply. But Penn nev­er men­tions that the Ben­dix­en poll had a mar­gin of error of plus-or-minus 4 per­cent–four being a larg­er num­ber than one. Addi­tion­al­ly, it’s mean­ing­less with­out fur­ther study. Any­one in the age brack­et would attribute it to video games, or snipers being, let’s admit it, quite cool. Yet Penn, based on no fol­low-up inter­views, detects a new patri­o­tism,” and a desire to mas­ter com­plex math­e­mat­i­cal for­mu­las like how dis­tance or wind might affect the path of the bul­let.” This sim­ply isn’t pro­fes­sion­al work. (It is bit­ter, though. Penn con­cludes the chap­ter by com­plain­ing, Ask any­one in pol­i­tics and they will agree – they face snipers’ every day who are try­ing to find one flinch, one out-of-place word to put on Drudge or YouTube.” It takes a spe­cial sort of self-regard to com­pare the dan­ger of being embed­ded on YouTube to being hunt­ed down in urban warfare.)

Else­where, Penn con­flates one poll on atti­tudes toward the reli­gion of Islam with atti­tudes toward Amer­i­can Mus­lims. At times, he mix­es per­cent­ages and absolute val­ues for scare effect, as when he dark­ly warns that if one-tenth of one per­cent of our pop­u­la­tion – 300,000 peo­ple – turned to al Qae­da, it would be more than enough to desta­bi­lize our soci­ety.” Some­times he just dis­cards data, as when he con­ve­nient­ly decides to ignore his evi­dence that church­go­ers reject female min­is­ters and spec­u­lates that con­sen­sus and com­pas­sion may be on the outs right now, but they are bound to make a come­back,” which is all the argu­ment he needs to say we are also ready for the first female Bil­ly Gra­ham.” And on, and on.

All this is in ser­vice of his con­cept that microtrends” now gov­ern our world: It takes 1 per­cent of peo­ple mak­ing a ded­i­cat­ed choice – con­trary to the main­stream’s choice – to cre­ate a move­ment that can change the world,” Penn writes. Why 1 per­cent? Who knows? Penn does­n’t stick to it him­self. Some­times, it’s one-tenth of one per­cent, as in his al Qae­da exam­ple, or 10 per­cent, as with left­ies, or some­times it’s the microtrend of – I kid you not – the tens of mil­lions of Amer­i­cans who moved to the sub­urbs in the 20th cen­tu­ry. Toward the book’s end, Penn says the mag­ic of the 1 per­cent thresh­old” is that ten peo­ple with bazookas can over­come 1,000 peo­ple with pick­et signs, but they can’t over­come 10,000 peo­ple with pick­et signs.” Chew on that one, grasshopper.

As microchap­ter after microchap­ter passed, review­ing this book began to feel like drop­ping a grenade into a bar­rel of fish. But Microtrends is illu­mi­nat­ing. Poll­sters occu­py a unique­ly pow­er­ful space in Amer­i­can polit­i­cal dis­course: They bring sci­ence to elec­tions. Armed with heaps of raw data, they ele­vate their opin­ions into some­thing alto­geth­er weight­i­er: Con­clu­sions. When an orga­ni­za­tion sends out a press release say­ing the orga­ni­za­tion is right, it’s ignored. When a poll­ster sends out a poll show­ing the elec­torate agrees, ears in Wash­ing­ton perk up.

The enter­prise has always been dodgy. Pop­ulist poll­sters reli­ably dis­cov­er that the elec­torate thirsts for more pop­ulism. Con­ser­v­a­tive poll­sters rou­tine­ly dis­cov­er a small gov­ern­ment con­sen­sus puls­ing at the heart of the body politic. When the lib­er­tar­i­an Cato Insti­tute com­mis­sioned a poll of the elec­torate, they found – shock­ing­ly – that the essen­tial swing vote was made of lib­er­tar­i­ans. Remark­ably, when­ev­er a politi­cian or self-inter­est­ed insti­tu­tion releas­es a poll, the results show a sym­me­try between the atti­tudes of the poll­ster’s employ­er and those of the vot­ers. But Pen­n’s book shines light on this phe­nom­e­non: If he is the pin­na­cle of his pro­fes­sion, then the pro­fes­sion uses num­bers as a ruse – a super­fi­cial empiri­cism that obscures gar­den-vari­ety hack­ery. And that’s a trend worth wor­ry­ing about.

Ezra Klein is a staff writer at the Amer­i­can Prospect who blogs reg­u­lar­ly at www​.ezrak​lein​.com.
Subscribe and Save 66%

Less than $1.67 an issue