The Secret Sinsof EconomicsDeirdre McCloskey

PRICKLY PARADIGM PRESSCHICAGO

Copyright © 2002 Prickly Paradigm Press, LLCAll rights reserved.

Prickly Paradigm Press, LLC5629 South University AvenueChicago, Il 60637

www.prickly-paradigm.com

ISBN: 0-9717575-3-4LCCN: 2002 102650

1

What’s sinful about economics is not what the averageanthropologist or historian or journalist thinks. Fromthe outside the dismal science seems obviously sinful, ifirritatingly influential. But the obvious sins are not allthat terrible; or, if terrible, they are committed anywayby everybody else. It is actually two particular, non-obvious, and unusual sins, two secret ones, that cripplethe scientific enterprise—in economics and in a fewother fields nowadays (like psychology and politicalscience and medical science and population biology).

Yet a sympathetic critic who says these things andwishes that her own beloved economics would grow upand start focusing all its energies on doing properscience (the way physics or geology or anthropology orhistory or certain parts of literary criticism do it) finds

we call English, certain Dutchmen) wanted most toknow How Much. It was an entirely novel obsession.You might call it bourgeois. How Much will it cost todrain the Somerset Levels? How Much does England’streasure by foreign trade depend on possessingcolonies? How Much is this and How Much that?The blessed Adam Smith a century later keptwondering How Much wages in Edinburgh differedfrom those in London (too much), and How Much thecolonies by then acquired in England’s incessant eigh-teenth-century wars against France were worth to thehome country (not much). By the late eighteenthcentury, it is surprising to note, the statistical chart hadbeen invented; what isn’t surprising is that it hadn’tbeen invented before—another sign that quantitativethinking was novel, at least in the West (the Chinesehad been collecting statistics on population and pricesfor centuries). European states from Sweden to Naplesbegan in the eighteenth century collecting statistics toworry about: prices, population, balances of trade, flowsof gold. The word “statistics” was a coinage ofGerman and Italian enthusiasts for state action in theearly eighteenth century, pointing to a story of the stateuse of numbering. Then dawned the age of statistics,and everything from drug incarcerations and smokingdeaths to the value of a life and the credit rating of JaneQ. Public are numbered.

It became a sort of insanity, of course. Tour guidesobserve that American men want to know how tallevery tower is, how many bricks there are in everynotable wall, how many died here, how many lived.

3

herself sadly misunderstood. The commonplace andvenial sins block scrutiny of the bizarre and mortalones. Pity the poor sympathetic critic, construed regu-larly to be making this or that Idiot’s Critique: “Oh, Isee. You’re one of those airy humanists who just can’tstand to think of numbers or mathematics.” Or, “Oh, Isee. When you say economics is ‘rhetorical’ you wanteconomists to write more warmly.”

I tell you it’s maddening. The sympathetic critic,herself an economist, even a Chicago-School econo-mist, slowly during twenty years of groping came torecognize the ubiquity of the Two Secret Sins ofEconomics (in the end they are one, deriving frompride, as all sins do). She has developed helpful sugges-tions for redeeming economics from sin. And yet noone—not the anthropologist or English professor orothers from the outside certainly, but least of all theeconomist or medical scientist—grasps her point, oracts on it.

VIRTUES MISIDENTIFIED AS SINS

Quantification

Quantification, though, is not a sin. Numbers camewith social science at its birth. The English politicalarithmeticians William Petty and Gregory King andthe rest in the late seventeenth century (anticipated inthe early seventeenth century by, like so much of what

2

it is no compliment to you to say, that you knowbetter. Now, what are the Facts of this case? Youare, we will say in round numbers, twenty years ofage; Mr. Bounderby is, we will say in roundnumbers, fifty....[T]he question arises, Is this onedisparity sufficient to operate as a bar to such amarriage? In considering this question, it is notunimportant to take into account the statistics ofmarriage, so far as they have yet been obtained, inEngland and Wales. I find, on reference to thefigures, that a large proportion of these marriagesare contracted between parties of very unequal ages,and that the elder of these contracting parties is, inrather more than three-fourths of these instances,the bridegroom. It is remarkable as showing thewide prevalence of this law, that among the nativesof the British possessions in India, also in a consid-erable part of China, and among the Calmucks ofTartary, the best means of computation yet furnishedus by travelers, yield similar results.”

Counting can surely be a nitwit’s, or the Devil’s, tool.Among the more unnerving exhibits in the extermina-tion camp at Auschwitz are the books in which Hitler’swilling executioners kept records on every person theykilled.

The formal and mathematical theory of statistics waslargely invented in the 1880s by eugenicists (thoseclever racists at the origin of so much in the socialsciences) and perfected in the twentieth century byagronomists (yes, agronomists, at places like theRothamsted agricultural experiment station in Englandor at Iowa State University). The newly mathematized

Samuel Johnson was in 1775 typical of his age and hisgender in reporting the size of everything he encoun-tered in his tour of the West of Scotland (he used hiswalking stick as a measuring rod). By the 1850s theconservative critics of capitalism, such as CharlesDickens, were becoming very cross about statistics:

Thomas Gradgrind, sir—peremptorilyThomas—Thomas Gradgrind. With a rule and apair of scales, and the multiplication table always inhis pocket, sir, ready to weigh and measure anyparcel of human nature, and tell you exactly what itcomes to. It is a mere question of figures, a case ofsimple arithmetic...

“Father,” she still pursued, “does Mr Bounderbyask me to love him?”

“...[T]he reply depends so materially, Louisa, onthe sense in which we use the expression. Now, MrBounderby does not do you the injustice, and doesnot do himself the injustice, of pretending toanything fanciful, fantastic, or (I am using synony-mous terms) sentimental....Therefore, perhaps theexpression itself—I merely suggest this to you, mydear—may be a little misplaced.”

“What would you advise me to use in its stead,father?”

“Why, my dear Louisa,” said Mr. Gradgrind,completely recovered by this time, “I would adviseyou (since you ask me) to consider this question, asyou have been accustomed to consider every otherquestion, simply as one of tangible Fact. The igno-rant and the giddy may embarrass such subjects withirrelevant fancies, and other absurdities that have noexistence, properly viewed—really no existence—but

54

It is a mere question of figures, a case of simple arith-metic.

Refutatio: But after all, think about it. When you wantto count your coconuts, or the cash value of yourChristmas gifts, it makes sense to do the job right.Many of the things we wish to know come in quantita-tive form. It matters—not absolutely, in God’s eyes,but for particular human purposes—how much it willrain tomorrow and how much it rained yesterday. Forsound practical and spiritual reasons we wish some-times to know How Much. How many slaves weredriven from Africa? Perhaps 29 million (the populationof Britain at the height of the slave trade was about 8million), more than half going east, not west, across theSahara or the Indian Ocean, not the Atlantic. How hasCuba fared under Communism and the Americanembargo? Income per head in Cuba has fallen by athird since 1959, while in the Dominican Republic,Chile, Mexico, Brazil, and indeed in Latin America andthe Caribbean generally it has more than doubled.How big is immigration to the United States now?Smaller in proportion to population than it was in1910. And on and on and on.

(You can see from the examples that no claim is beingmade here that numbers are by nature peculiarly “objec-tive,” whatever that pop-philosophical term mightexactly mean, or “non-political,” or “scientific.”Numbers are rhetoric, which is to say humanly persua-sive. We agree in a persuasive culture to assignmeaning to this or that number, and then can be

statistics became a fetish in wannabe sciences. Duringthe 1920s, when sociology was a young science, quan-tification was a way of claiming status, as it became alsoin economics, fresh from putting aside its old name ofpolitical economy, and in psychology, fresh from aseparation from philosophy. In the 1920s and 1930seven the social anthropologists, those men and womenof the fanciful, fantastic, or (I am using synonymousterms) sentimental, counted coconuts.

And the economists, oh, the economists, how theycounted, and still count. Take any copy of TheAmerican Economic Review to hand (surely yousubscribe?) and open it at random. To perhaps JoelWaldfogel, “The Deadweight Loss of Christmas” (nokidding: December 1993; Waldfogel is arguing thatsince a gift is not chosen by the recipient it is not worthwhat the giver spent, which leads to a loss comparedwith merely sending cash. Who could not love such ascience of Prudence?). On p. 1331 you will find thefollowing Table 1:

Average Amounts Paid and Values of Gifts

Variable Survey 1 Survey 2Amount paid ($) 438.2 508.9Value ($) 313.4 462.1Percentage ratio ofaverage value to 71.5 90.8average paidNumber of recipients 86 58

76

Mathematics

Nor is mathematics a sin. Mathematics is not identicalto counting or statistics. The newspapers chortlewhen they find a mathematician who cannot balancehis checkbook, but that’s just a misunderstanding ofwhat mathematicians do. There have been somefamously good calculators among mathematicians, the eighteenth-century Swiss mathematician LeonhardEuler being an instance (he also knew the entire Aeneidby heart; in Latin, I need hardly add). But odd as itsounds, most of mathematics has nothing to do withactual numbers. Euler used calculation in the same waythat mathematicians nowadays use computers, forback-of-the-envelope tests of hunches on the way todeveloping what the mathematicians are pleased to calla real proof of such amazing facts as: eπi + 1 = 0 (andtherefore God exists). You can have a “real” proof, thestyle of demonstration developed by the Greeks (withwhich you became acquainted in high-school geom-etry, either loving or hating it), without examining asingle number or even a single concrete example.Thus: the Pythagorean Theorem is true for any righttriangle, regardless of its dimensions, and is proven notby induction from many or even zillions of numericalexamples of right triangles, but universally and for alltime, praise God, may her name be glorified, bydeduction from premises. Accept the premises andyou have accepted the Theorem. Quod erat demon-strandum.

persuaded to this or that view of the matter. Pebbles liearound, as Richard Rorty has put it; facts of the matterdo not. It is our human decision to count or weigh ormix the pebbles in constituting the pebbly facts.)

Economists are selected for their great love of numbers.The joke is, “I’m an economist because I didn’t haveenough personality to become an accountant.” A statis-tical argument is always honored in the Department ofEconomics. Many non-economists on the contrary fearnumbers, dislike them, dishonor them, are confused andirritated by them. But some important questions canonly be answered numerically. A great many otherquestions are at least helpfully illuminated by numbers.Your age number is not the only important fact aboutyou, and is certainly nothing like your full Meaning(“You are, we will say in round numbers, twenty years ofage; Mr Bounderby is, we will say in round numbers,fifty”). But it is a number helpful for some purposes—ordinary conversation, for one thing; medical examina-tion for another; yes, even marriage. It’s humanly usefulto know that you grew up in the 1950s and came of agein the liberating 1960s: age sixty on September 11, 2002(happy birthday). Temperature is not the only measureof a good day. Wind, sunshine, human events, andhuman-assigned significance matter. That this is themonth and this the happy morn of Christ’s nativity hasmeaning beyond 30ºF. But it is worth knowing that thetemperature on the blessed day was not -459.67ºF or212ºF.

So counting is not a sin of economics. It is a virtue.

98

Much, the quantitative effect, the magnitude, the mass,the oomph.

Economics since its beginning has been very often“mathematical” in this sense of being interested inWhy/Whether arguments without regard to How Much.For example: If you buy a loaf of bread from the super-market both you and the supermarket (its shareholders,its employees, its bread suppliers) are made to somedegree better off. How do I know? Because the super-market offered the bread voluntarily and you acceptedthe offer voluntarily. Both of you must have been madebetter off, a little or a lot—or else you two wouldn’thave done the deal.

Economists have long been in love with this simpleargument. They have since the eighteenth centurytaken the argument a crucial and dramatic step further:that is, they have deduced something from it, namely,Free trade is neat. If each deal between you and thesupermarket, and the supermarket and Smith, andSmith and Jones, and so forth is betterment-producing(a little or a lot: we’re not talking quantities here), then(note the “then”: we’re talking deduction here) freetrade between the entire body of French people and theentire body of English people is betterment-producing,too. And therefore (note the “therefore”) free tradebetween any two groups is neat. The economist notesthat if all trades are voluntary they all have some gain.So free trade in all its forms is neat. For example, a lawrestricting who can get into the pharmacy business is abad idea, not neat at all, because free trade is good, so

1110

Statistics or other quantitative methods in science (suchas accounting or experiment or simulation) answerinductively How Much. Mathematics by contrastanswers deductively Why, and in a refined and philo-sophical version very popular among mathematicianssince the early nineteenth century, Whether. “Whydoes a stone dropped from a tower go faster andfaster?” Well, F = ma, understand? “I wonder Whetherthe mass, m, of the stone has any effect at all.” Well,yes, actually it does: notice that there’s a little m in theanswer to the Why question.

Why/Whether is not the same question as How Much. Youcan know that forgetting your lover’s birthday will havesome effect on your relationship (Whether), and evenunderstand that the neglect works through such-and-such an understandable psychological mechanism(“Don’t you love me enough to know I care about birth-days?”—Why). But to know How Much the neglectwill hurt the relationship you need to have in effectnumbers, those ms and as, so to speak, and some notionof their magnitudes. Even if you know the Why (theproper theory of the channels through which forgettinga birthday will work; again by analogy, F = ma), theHow Much will depend on exactly, numerically, quanti-tatively how sensitive this or that part of the Why is infact in your actual beloved’s soul: how much in this casethe m and a are. And such sensitivity in an actualworld, the scientists are always saying, is an empiricalquestion, not theoretical. “All right, you jerk, that’s thelast straw: I’m moving out” or “Don’t worry, dear: Iknow you love me” differ in the sensitivity, the How

that any individual exchange arrived at voluntarily isgood, then with a few extra assumptions (e.g., about themeaning of “voluntarily”; or, e.g., about how oneperson’s good depends on another’s) you can get theconclusion that free international trade among nationsis good.

Why/Whether reasoning, which is also characteristic ofthe Math Department, could be called philosophical.The Math Department and the PhilosophyDepartment have a similar fascination with deduction,and a corresponding boredom with induction. Theydo not give a fig for How Much. No facts, please:we’re philosophers. No numbers, please: we’re mathe-maticians. In the Philosophy Department either rela-tivism is or is not open to a refutation from self-contra-diction. It’s not a little refuted. It’s knocked down, ornot. In the Math Department the GoldbachConjecture, that every even number is the sum of twoprime numbers (e.g., 24 = 13 + 11; try it), is either trueor false (or, to introduce a third possibility admittedsince the 1930s, undecidable). Supposing it’s decidable,there’s no question of How Much. You can’t, in therealm of Why/Whether, in the Math Department orthe Philosophy Department or some parts of theEconomics Department, be a little bit pregnant.

The argument for free trade is easy to express in termsthat anyone would call “mathematical.” Since about1947 the front line and later the dominant and by nowthe arrogantly self-satisfied and haughtily intolerant ifremarkably unproductive scientific program in

non-free trade is bad. Protection of French workers isbad, because free trade is good. And so forth, to liter-ally thousands of policy conclusions.

Though it is among the three or four most importantarguments in economics, it is not empirical. It containsno statements of How Much. It says there exists a gainfrom trade—remember the phrases some gain or to somedegree or a little or a lot or we’re not talking quantitieshere. “I wonder Whether there exists [in whateverquantity] a good effect of free trade.” Yes, one exists:examine this page of math; look at this diagram; listento my charming parable about you and the super-market. Don’t ask How Much. The reasoning isWhy/Whether. As stated it cannot be wrong, no morethan the Pythagorean Theorem can be. It’s not amatter of approximation, not a matter of How Much.It’s a chain of logic from implicit axioms (which can beand have been made explicit, in all their infinite variety)to a “rigorous” qualitative conclusion (in its infinitevariety). Remember those words “then,” “therefore,”“so.” Under such-and-such a set of assumptions, A, theconclusion, C, must be that people are made better off.A implies C, so free trade is beneficial anywhere.(Please listen, and stop asking “How Much?”: how manytimes must I remind you that the reasoning is qualita-tive, not quantitative?!)

The philosophers call this sort of thing “valid”reasoning, by which they do not mean “true,” but“following from the axioms—if you believe the axioms,such as A, then C also must be true.” If you believe

1312

Why/Whether arguments about economics).Deducing sometimes surprising and anyway logicallyvalid (if not always true) conclusions from assumptionsabout the economy is a game economists have alwaysloved. And if you want to connect one thing withanother, deduce conclusions C from assumptions A,free trade from characterizations of an autonomousconsumer, why not do it universally and for all time?Why not, asked Samuelson and Arrow and the rest,with much justice, do it right?

True, for practical purposes of surveying grain fields itwould work just as well as Pythogoras’ Greek proof tohave a Babylonian-style of proof-by-calculationshowing that the sums of squares of the sides of zillionsof triangles seem to be pretty much equal to the sumsof squares of their hypotenuses. You might make asimilar case for the free trade theorem, noting forexample that the great internal free-trade zone calledthe United States still has a much higher averageincome (20 to 30 percent higher) than otherwise cleverand hard working countries like Japan or Germany,which insist on many more restrictions on internaltrade, such as protection of small retailing. And, true,the improvement of computers is making moreBabylonian-style “brute force calculations” (as themathematicians call them with distaste) cheaper thansome elegant formulas (“analytic solutions,” they say,rapturously). Economics, like many other fields—architecture, engineering—is about to be revolution-ized by computation.

economics has been to reformulate verbal (but stillphilosophical/mathematical, i.e. qualitative, i.e.Why/Whether) arguments into symbols and variablesand diagrams and fixed point theorems and the like.The program is called “Samuelsonian,” after the Gary,Indiana native and third person to receive the NobelMemorial Prize in Economics, Paul AnthonySamuelson. He and his brother-in-law Kenneth Arrow(who was the fifth person out of the fifty or so from1969 to 2001 to receive the glittering Prize) led themovement to be explicit about the math in economics,against great opposition. They were courageouspioneers (their mutual nephew Lawrence Summers, thecrown prince of modern economics, became Secretaryof the Treasury and President of Harvard). In 1947Samuelson set the tone with the publication of hisPh.D. dissertation (which had been finished in 1941),the modestly entitled Foundations of Economic Analysis.In 1951 Arrow carried it to still higher realms of math-ematics with his Ph.D. dissertation, Social Choice andIndividual Values. Their enemies, a few of whom arestill around, said, with the humanists, “Yuk. This mathstuff is too hard, too inhuman. Give me words.Sentiment. Show me some verbal argumentation orsome verbal history. Or even actual numbers. Butnone of this new x and y stuff. It gives me a headache.”

Refutatio: But think again. There’s nothing whatevernew about deductive reasoning in economics. It didn’tstart in 1947. More like 1747 (in fact about this timeDavid Hume in Scotland and the physiocrats in Francewere busy inventing philosophical, entirely qualitative,

1514

usually been libertarian. Economists are freedom nuts,which is to say that they look with suspicion onlawyerly plans to solve problems with new statecompulsions and longer jail sentences. Economics atits philosophical birth, among physiocrats in Paris andmoral philosophers in Edinburgh, was in favor of freemarkets and was suspicious of overblown states.Mostly it still is. Let things be, laissez faire, has beenthe economists’ cry against intervention. Let the tradesbegin.

True, not all economists are free traders. The non-freetraders, often European and disproportionately nowa-days French, point out that you can make otherassumptions about how trade works, A', and get otherconclusions, C', not so favorable to laissez faire. Thefree-trade theorem, which sounds so grand, is actuallypretty easy to overturn. Suppose a big part of theeconomy—say the household—is, as the economistsput it, “distorted” (e.g., suppose people in householdsdo things for love: you can see that the economists havea somewhat peculiar idea of “distortion”). Then itfollows rigorously (that is to say, mathematically) thatfree trade in other sectors (e.g., manufacturing) will notbe the best thing. In fact it can make the averageperson worse off than restricted, protected, tariffedtrade would.

And of course normal people—I mean non-econo-mists—are not persuaded that free trade is always andeverywhere a good thing. For example most peoplethink free trade is a bad thing for the product or service

But if beyond clumsy fact or numerical approximationthere is an elegant and exact formula—F = ma or E =mC2 or, to give a somewhat less elegant example fromeconomics, 1 + iusa = (eforward / espot) (1 + ifrance), called“covered interest arbitrage”—why not use it? Ofcourse, any deduction depends on the validity of thepremises. If a sufficiently high percentage of potentialarbitrageurs in the markets for French and U.S. bondsand currency are slothful dolts, then covered interestarbitrage will not hold. But likewise any inductiondepends on the validity of the data. If the sample usedto test the efficacy of mammograms in preventingpremature death is biased, then the statistical conclu-sions will not hold. Any calculation depends on thevalidity of the inputs and assumptions. Garbage in,garbage out. As the kids say, it all depends. Naturally:we mortals are not blessed with certitude.

So mathematics, too, is not the sin of economics, but initself a virtue. Getting deductions right is the Lord’swork, if not the only work the Lord favors. Like allvirtues it can be carried too far, and be unbalanced withother virtues, becoming the Devil’s work, sin. But allvirtues are like that.

Libertarian Politics

Nor is devotion to free markets a sin. Like quantitativeinduction and philosophical deduction, economics hasalways had a political purpose, and the purpose has

1716

think, think again. There really is a serious case to bemade against government intervention and in favor ofmarkets. Maybe not knockdown; maybe imperfect hereor there; let’s chat about it; hmm, I see what you mean;but a serious case that serious people ought to takeseriously. The case is not merely Country-ClubRepublicanism (which in fact is highly favorable togovernment intervention, in order you see to assist themembers of the Country Club, such as its longstandingmembers who managed Enron, Inc.). The case in favorof markets is on the contrary populist and egalitarianand person-respecting and bad-institution-breakinglibertarianism. Don’t go to government to solve prob-lems, said Adam Smith. As he didn’t say, to do so is toput the fox in charge of the hen house. The goldenrule is, those who have the gold rule: so don’t expect agovernment run by men to help women, or a govern-ment run by Enron executives to help Enronemployees.

Libertarianism is typical of economics, especiallyEnglish-speaking economics, and most especiallyAmerican economics. Most Americans if they can getclear of certain European errors, are radical libertariansunder the skin. Give me liberty. Sweet land of liberty.Live free or die (a New Hampshire man who decidedhe didn’t want the motto on his license plate andinsisted on covering it up with masking tape was...arrested: your friend the State in action).

But alas, no time, no time. Libraries of books havebeen written examining the numerous and weighty

they make. By all means, let us arrange for my bakerand pharmacist to compete vigorously, nay, brutally,with other bakers and pharmacists, so that I can getdonuts and also vitamin E (to offset the donuts)cheaply. But I really do think we need to blockadeentry into the profession of being an economist: it is, Iam sure you agree, scandalous that so many unqualifiedquacks are bilking consumers with adulteratedeconomics, quite unlike the pure economic ideas I offerhere, at such reasonable expense.

And very many normal people of leftish views, evenafter communism, even after numerous disastrousexperiments in central planning, even after trying toget a train ride from Amtrak or service from the PostalService (not to mention service from the InternalRevenue Service or from the Immigration andNaturalization Service; you see I wax indignant: I am,after all, a free-market economist), think SocialismDeserves a Chance. They think it obvious thatsocialism is after all fairer than unfettered capitalism.They think it obvious that regulation is after all neces-sary to restrain monopoly. They don’t realize that freemarkets have partially broken down inequality (forexample, between men and women; “partially,” I said)and partially undermined monopolies (for example,local monopolies in retailing) and have increased theincome of the poor over two centuries by a factor of18. The sin of economics, the lefties think, is exactlyits free-market bias.

Refutatio: But, my dearly beloved friends on the left,

1918

the forms filled out and visiting the right governmentoffices recently took a team of researchers working sixhours a day 289 days. To get the permits to build alegal house on state-owned land (land for sale, not heldfor the public) took nearly seven years, with 207 admin-istrative steps and 52 government offices. In Egyptgetting the permits to build a legal house on agricul-tural land took from 6 to 11 years. In Haiti buying landfrom the government took 19 years.

Nor is such government obstruction peculiar to thepresent-day Third World. In one decade in the eigh-teenth century, according to the Swedish economistand historian Eli Heckscher in his book of 1932,Mercantilism, the French government sent tens of thou-sands of souls to the galleys and executed 16,000 (that’sabout 4.4 people a day over the ten years: you see thebeauty of statistical thinking) for the hideous crime of...are you ready to hear the appalling evil these enemiesof the State committed, fully justifying hanging themall, every damned one of their treasonable skins?...importing printed calico cloth. States do not changemuch from age to age. Lawrence Wylie reported theattitude of a French bureaucrat in the 1950s: “If thepublic speaks evil of me I serenely shit on it. Thecomplaint merely goes to show the value of my officeand of my methods. The more the public is shat upon,the better the State is served.”

In view of How Muches and Oh, My Gods like these—the baleful oomph of governmental intrusions world-wide crushing harmless (indeed, beneficial) exchange,

arguments for the market and against socialism. I urgeyou to go read a few such books with care, such asThomas Friedman, The Lexus and the Olive Tree, or ifyour tastes run more academic, anything by MiltonFriedman (Nobel 1976). Please, all of you, come overto my delightful, if challenging, course at theUniversity of Illinois - Chicago called “Economics forAdvanced Students of the Humanities” in which Isketch the arguments. Really, that the average literaryperson believes the first few pages of The CommunistManifesto suffice for knowledge of economics andeconomic history, in which he professes great interest,is a bit of a scandal. It’s amazing that most professorsand journalists since about 1900 have not even heard ofthe arguments against turning the economy over topolice and jailers and bureaucrats, and are scandalizedwhen some boorish Chicago-School economist comesalong and suggests that pot should be legalized andnational borders opened and government schools madeto compete with each other. I spoiled quite a fewdinner parties early in my career blurting out suchproposals. I have become cannier since then, or morepolite, or just weary.

But I say, as Cromwell said wearily to the GeneralAssembly of the Church of Scotland, 3 August, 1650, “Ibeseech you, in the bowels of Christ, think it possibleyou may be mistaken.”

Oh, permit me one short libertarian riff. According tothe Peruvian development economist Hernando deSoto, to open a small business in Lima, Peru getting all

2120

Paul Samuelson (b. 1915), who fully formalized thenotion in a curious character known as Max U, and thegreat Gary Becker (b. 1930), who went about as far ashe could go.Becker (Nobel 1992), a professor of economics andsociology at the University of Chicago, asks, forexample, why people have children. Answer: becausechildren are durable goods. They are expensive toproduce and maintain, over a long period of time, like ahouse. They yield returns over a long future, like a car.They have a poor second-hand market, like a refriger-ator. They act as a store of value against future disas-ters, like pawnable gold or your diamond ring. So (youwill sense a logical leap here; David Hume noticed thesame leap in Mandeville and Hobbes), the number ofchildren that people have is a matter of cost andbenefit, just like the purchase of a house or car orrefrigerator or diamond. A prudent parent decideswhether to invest in many children or few, extensivelyor intensively, early or late, just like investing in adurable good.

If you think this is funny stuff you are not alone. Butthink again: there’s no doubt that Prudence does affectat least part of the decision to have children, toemigrate, to attend church, to go to college, to commita murder, not to speak of buying a house or a car or aloaf of bread. In his obsessive study of the Prudentialpart, the economist can make some quite interestingand sometimes counter-intuitive and occasionally evenfactually true points. For example, economists“predict” (as they always put it in their child’s version

from marijuana to printed calico—perhaps laissez fairedoes not seem so obviously sinful, does it now?Consider, my dear leftist friends. Read and reflect. Ibeseech you, think it possible that, like statistics andmathematics, the libertarianism of economics is avirtue.

VENIAL SINS, EASILY FORGIVEN

I am very far from wanting to defend everything abouteconomics, even short of the Two Great Secret Sins.But you need to realize that economists do the irritatingthings they do for reasons, often pretty good ones.

For instance, among the most surprising and irritatingfeatures of economics (when people figure out what isgoing on) is its obsessive, monomaniacal focus on aPrudent model of humanity. It’s hard for outsiders tobelieve. Everything, simply everything, from marriageto murder is supposed by the modern economist to beexplainable as a sort of Prudence. Human beings aresupposed to be calculating machines pursuing Prudenceand Price and Profit and Property and Power—“P vari-ables,” you might call them. P-obsession begins withMachiavelli and Hobbes, is continued by BernardMandeville (the early eighteenth-century Dutch-English spy and pamphleteer), is systematized byJeremy Bentham (the utilitarian economist flourishingin the early nineteenth century), and is finally perfectedby twentieth-century economists, including that same

2322

Only men from Machiavelli to Becker are claiming isthat you can explain B with Prudence alone, the P vari-able—Prudence, Price, Profit, The Profane. Smith(and Mill and Keynes and quite a few other economists,if not the ones who run the discipline these days) havereplied that, no, you have forgotten Love and Courage,Justice and Temperance, Faith and Hope, in a word,Solidarity, the S variable of speech, stories, shame, TheSacred. Economists have specialized in P, anthropolo-gists in S. But most behavior, B, is explained by both:

B = a + ßP + yS + E.

To include both P and S is only sensible. It is notwishy-washy or unprincipled. Of course the S vari-ables are the conditions under which the P variableswork, and of course the P variables modify the effectsof S variables. It is the human dance of Sacred andProfane.

(Econometrically speaking, I remind my economistcolleagues, if the P and S variables are not orthogonal,which is to say if they are not entirely independent, orthe covariance, as we say, of P and S is not zero, byGod’s grace, bless her holy name, or alternatively ifthere is reason to believe that a variable such as PSmultiplied together (say) has its own influence, then anestimate of the coefficients a and ß that ignores S (orPS) will give biased results. The bias is important ifthe S variable is important. The experiment is notproperly controlled, and its conclusions are nonsense.)

of positivism) that, surprisingly, no-fault divorce shouldhave no long-term effect on the prevalence of divorce.Why is that? Well, the law affects how the spoils froma divorce are divided up, but not their total size. Sincethe people on both sides have lawyers paid to collectspoils, it is the sum of spoils, not their division, whichshould in fact determine how much divorcing goes on.That the wife gets half instead of one quarter is offsetby the necessary concomitant: the husband thereforegets half instead of three-quarters. Her increasingpropensity to seek divorce (half is better than onequarter) is offset by his decreasing propensity (a half isworse than three quarters). And such a surprisingclaim on the basis of Prudence alone seems to be factu-ally true in the world.

The narrowness of the scientific concern of economistshas of course a cost (which is itself an economist’spoint: the road not traveled is the opportunity cost). Prudence is the central ethical virtue of the bour-geoisie, but not the only one. Adam Smith’s bookabout Prudence, An Inquiry into the Nature and Causes ofthe Wealth of Nations, published in 1776, should be readas embedded in the other virtues, especiallyTemperance and Justice, about which indeed Smithwrote at great length. If Smith had been statisticallyinclined then he would have put it this way. Take anysort of behavior you wish to understand—voting, forexample, or the adoption of the Bessemer process inthe making of steel. Call it B. It can be put on a scaleand measured, or perhaps seen to be present or absent.You want to give an account of B. What the Prudence-

2524

coming by age 50 or so to realize that, after all, peopleare motivated by more than Prudence. Even GaryBecker shows signs of such a development.

To this the academic economist who has not developedbeyond his graduate-student version of the science islikely to reply, following the P-Only model, “Thanksfor the advice. But I make a good living specializing inP variables.” His sin is a selfish species of Ivory-Towerism. “Why do I need to concern myself with theentire argument? I do my specialty.”

Well, so what? Don’t you want to get the correctanswer; or do you merely want to collect yourpaycheck? (Don’t answer that.)

NUMEROUS WEIGHTY SINSREQUIRING SPECIAL GRACE TOFORGIVE BUT SINS NOTPECULIAR TO ECONOMICS

And then there are sins less easily forgiven, less easilyput down to a prudent specialization that at least keepsP variables in the scientific game. The sins areshameful and scientifically damaging, I admit, havingmyself committed all of them at one time or another,sometimes for years and years. I am truly sorry and Ihumbly repent. But, goodness, if you are going to

It is often a mistake to rely on S alone, and to reject P,as Marshall Sahlins sometimes seems to do (shame onyou, Marshall; he says he doesn’t, but I say he does).And vice versa, which is the point here. Mosteconomics and most anthropology is persuasion aboutthe mixture of Prudence and Solidarity, the Profaneand the Sacred, that matters for any particular case.Without being explicit enough, some economists, andsome of the best, do acknowledge S variables.Theodore Schultz argued in Transforming TraditionalAgriculture (1964; Nobel 1979) that peasants in poorcountries were Prudent. He was arguing that it was amistake to explain their behavior anthropology-style asB = a + yS + E with the S variable alone. Schultz said:Even these “traditional” peasants care about P. ButSchultz did not ignore the S variables. The educationof women, he argued forcefully, was crucial in makingPrudence work, and doing it would depend on over-coming patriarchal objections to literate women.Robert Fogel (Nobel 1993) and Stanley Engermanargued in 1974 that American slavery was Prudentialand capitalistic. But they did not entirely ignore the Svariables. They measured them, by indirection,finding that for some features of slavery, such as theprice of slaves, variables other than business Prudencewere quantitatively not very important. And thenFogel went on to write about the influence of religiousbelief on slavery and abolition, and Engerman to writeabout the historical roots of coercion and freedom inthe labor market. Many economists go through aBildung of this sort, starting in graduate school asPrudence-Only guys (the guys more than the gals) and

2726

About three-and-a-half percent.

The figure was so shocking even to economists that itbecame part of an investigation into graduateprograms by the American Economic Association.Reform was blocked by a member of the committee,also at the University of Chicago (are you seeing apattern here?), who wants the math-with-Prudence-Only game to go on and on, undisturbed by scientificconsiderations.

Outsiders would likewise be amazed at the HistoricalIgnorance of the economist. They think that the scien-tific evidence about economies before the past fewyears would surely figure in an economist’s data. Itdoesn’t. One graduate program after another in the1970s and 1980s cut the requirement that studentsbecome familiar with the economic past. I myselfmanaged for twelve years to fend off the day of execu-tion at the University of Chicago (now do you see thepattern?). The very month I left the department indisgust the barbarians inside the gates sent theeconomic history requirement to the guillotine, andsince then Ph.D.s in economics from the University ofChicago have joined those at Minnesota, Princeton,and Columbia in ignorance of the economic past. Atthe same time almost all American graduate programs(my own fair Harvard was proudly among the first todo so) were abandoning the study of the past ofeconomics itself. People call themselves economistswho have never read a page of Adam Smith or KarlMarx or John Maynard Keynes. It would be like being

damn economics for these you are going to have to lineup for damnation a considerable portion of the intelli-gentsia, commencing probably with your own sweetself.Economists, for example, are Institutionally Ignorant,which is to say that they don’t have much curiosityabout the world they are trying to explain. Forexample—this will surprise you—academic economists,especially since Samuelsonianism took over, have cometo think it is simply irrelevant, a waste of time, to doactual field work in the businesses they talk about.This is because (as they will explain to you patiently)people might lie, a point which is taken among econo-mists to be a profound remark in proper scientificmethod. So (you will see the non sequitur) never ask abusinessperson why she does something. Just observe,as though people were ants. The great economistRonald Coase (also at the University of Chicago,Nobel 1991, but taking a different approach to P and Sthan Gary Becker does—Coase is no Samuelsonian),while still a student at the London School ofEconomics, had the startling idea of actually speaking tobusinesspeople. He has been trying ever since aboutOctober 1932 to get other economists to do the samething. No soap. When two economists, Arjo Klamerand David Colander, asked economics graduatestudents what the skills were that made for a goodeconomist, nearly two thirds named mathematicalability and the ability to think up quick little models ofPrudence Only. How many named knowledge of theeconomic world as important? Go ahead, guess.

2928

ciple (economists have themselves stumbled on analo-gous findings in their own highly non-humanistic work,such as the finding of “rational expectations” or “thecheap talk paradox”). A famous story in linguisticsillustrates the point. A very pompous linguist wasgiving a talk at Columbia and noted that there werelanguages in which a double negative meant a positive(standard English, for example: “I am not going to notspeak” = “I am going to speak”) and languages in whicha double negative is a stronger negative (standardFrench and Italian, for example; or non-standardEnglish: “You ain’t got no class”). But, says he, articu-lating what he imagined was a universal of grammar,“There are no languages in which a double positive is anegative.” Pause. Silence. Then came a loud andknowing sneer from the back of the room: “Yeah,yeah.”

Their high-school version of positivism means theeconomists depend on a high-school version of the philos-ophy of science. “Well, you see: if H implies O, then itfollows rigorously that not-O implies not-H. So I canfalsify a hypothesis simply by looking at the observableimplications, O. What a wonderful simplification ofmy obligation to make scientific arguments! I can testthe hypothesis that people vote their pocketbooks, forexample, just by looking at how a party’s platformwould affect voter Smith or Jones in their pocketbooks.And if it’s not so falsified, it’s confirmed, right?”

Never mind that Pierre Duhem pointed out as longago as 1906 that the argument is nonsense in actual

3130

an anthropologist who had never heard of Malinowskior an evolutionary biologist who had never heard ofDarwin.

The more general Cultural Barbarism of economists iswell illustrated by their Philosophical Naïveté. Few econ-omists read outside economics. It is unnerving to gazeabout the library of a distinguished professor ofeconomics and find no books at all except on appliedmath and statistics: these are the worldly philosopherswho run our nation? Uh-oh. So naturally the profes-sors of economics have childish ideas about, say, episte-mology. They think for example that early logical posi-tivism (c. 1920), misunderstood because received thirdor fourth hand, is the latest philosophical word onmeaningfulness. “Let’s see now: I think I can recallfrom my high school physics course. If a Hypothesis,H, does not imply materially observable Observations,O, then it is ‘meaningless,’ right? So that means...‘means’? Uh... well, let it go... that all ethics, intro-spections, accounts of mental states, metaphors, framesof meaning, literature and myths—and it would seemall of mathematics and philosophy itself, I guess; butthat can’t be right—are meaningless blabber. Hmm.There must be something wrong here. Well, goodenough for government work.”

The economists know nothing of the main finding oflinguistics, philosophy, and literary criticism in thetwentieth century, namely, that we have ways of worldmaking, language games, senses of an ending thatcannot be reduced to formal grammars, even in prin-

(hmm...) stick to the positive. I know it’s hard tobelieve, but most economists really do think that thepositive/normative distinction lets them out of anyreflection on ethics. They want to believe that:“Economics is like astronomy in having nothing to dowith human affairs and therefore with the ethicaluniverse in which humans live. No, wait, that can’t beright: it has to do with human affairs—how else am Igoing to get paid for consulting or editorializing?—butthe parts I deal with are Objective... like who gets hurtby the imposition of free trade. Hmm. I’m havingtrouble with this. What I’m sure is that ‘ought’ and ‘is’are entirely different realms and the scientist ought toignore... uhm... well....”

And economists are tempted to arrogance in social engi-neering. Most humanists do not face the problem, sincepoets seldom think to ask English professors how towrite poems—though of course “criticism” in thebelletristic, three-star-awarding, judgments-of-Greatness sense does face the temptation, and normallyyields to it; and in fact many poets have been influencedby criticism (Poe’s criticism inspired Baudelaire;Emerson’s inspired Whitman). Anthropologists knowabout the problem in their own work, and worry: am Ibecoming a tool of Western imperialism?

Since economists think themselves well informedabout ethical philosophy if they have a muddy under-standing of positive vs. normative, you can imaginethe results. I would not want to accuse my colleaguesof being engineers devising efficiently operating exter-

science because every experiment or observation hasscientific controls (for example, S variables; ormeasuring devices and measuring errors) the truth orirrelevance of which needs to be assumed to make thetest work. (Economists call this the specificationproblem.) So the specification is actually H and S1 andS2 and S3 and... implies not-O or not-S1 or not-S2 or not-S3 or ....This means that the “falsifying” observationmay actually be a result of some failure of experimentalcontrol. And in fact on the frontiers of science themost usual quarrels are about just such matters: haveyou failed to control properly? Is your specificationright? Is it rational to expect people to be rational in avoting booth when they have already shown their irra-tionality by showing up at the polls in the first place,considering that their (or rather, his or her) single voteis virtually certain not to change the outcome? Haveyou properly controlled for social solidarity and senti-ment and other S variables affecting the vote, and arethese uncorrelated with the included variable, P, thepocketbook effect?

The words “metaphysical” or “philosophical” are usedin economics nowadays as terms of contempt: “That’srather philosophical, isn’t it?” means, “What a stupid,unscientific point; only an English professor would saysuch a thing!” So not surprisingly economists adherewithout criticism to, for instance, a high-school version ofethical philosophy. Economists believe that scientific andethical questions are distinct, the one “positive” and theother “normative,” and that real scientists ought to

3332

in this connection are: Paul Krugman (gold medal inthis category), Robert Lucas (Nobel 1995), andDeirdre McCloskey (bronze). Lots of intellectualprofessions are arrogant. Physicists, for example, arecontemptuous of chemists, whom they regard asimperfect versions of themselves. In fact physicists arecontemptuous of most people. But when a physicist atNorth Carolina named Robert Palmer went in 1989 toa conference in which physicists and economists wereto educate each other he remarked, “I used to thinkthat physicists were the most arrogant people in theworld. The economists were, if anything, more arro-gant.” I’m afraid he’s right on this score. Though ofcourse in general he’s a dope: a mere physicist.

Apologia: I have not, I realize, painted a very attractivepicture of economics. But these sins are widespread, Irepeat, among non-economists, too—even that oddone, candid selfishness, which you can find Nature’sEconomists articulating even when they aren’t trainedin it. But I earnestly invite you to learn by furtherreading in the literature the offsetting merits of econo-mists:

Economists are for one thing serious about the publicinterest, and are often the only people defending itwith any sort of lucidity and persuasiveness against thespecial interests. The model of worldly philosophywas originated in crude form by the early pamphle-teers and political arithmeticians (among them DanielDefoe). Adam Smith a half century and more laterbrought it to perfection.

mination chambers. At least not often. The liber-tarian streak in economics sometimes stays their hand.An economist would not view poor people as cattle tobe herded into high-rise concentration camps—asarchitects in the 1950s, for example, demonstrablydid; and as D. H. Lawrence and other democracy-haters earlier did. Or would they? What ethicalconsideration would stop them?

And economists are prone to an odd personalitydefect arising from their P-Only models, candid selfish-ness. When you ask a Chicago-School economist,“George, would you cooperate on this?” he is liable toanswer, “No: it’s not in my self-interest: don’t youbelieve in economics?” When I left Chicago so longago one of these people came up to me and said, “Isuppose you aren’t going to help grade the core exam-ination—after all, you’re out of here.” I was aston-ished, and replied, “No, I’m going to fulfill myremaining obligations.” He in turn was astonished. Ido not think it raised his opinion of me, that I was soinconsistent in advocating a P-Only theory ineconomic history (as I was then) but not in everydaylife. You mean you don’t cheat your employer whenyou get a chance? You mean you don’t imposeburdens on your colleagues when it serves yournarrow interests? Huh? What kind of an economistare you?

And I have to mention finally the very widespreadopinion that economists are prone to the sin ofpride—personal arrogance. Some names that come up

3534

THE TWO REAL SINS, ALMOSTPECULIAR TO ECONOMICS

A real science, or any intelligent inquiry into the world,whether the study of earthquakes or the study ofpoetry, economics or physics, history or anthropology,art history or organic chemistry, a systematic inquiryinto one’s lover or a systematic inquiry into the Dutchlanguage, must do two things. If it only does one ofthem it is not an inquiry into the world. It may begood in some other way, but not in the double way thatwe associate with good science or other good inquiriesinto the world, such as a detective solving a case.

I am sure you will agree: An inquiry into the worldmust think and it must look. It must theorize and mustobserve. Formalize and record. Both. That’s obviousand elementary. Not everyone involved in a collectiveintelligent inquiry into the world need do both: thedetective can assign his dim-witted assistant to justobserve. But the inquiry as a whole must reflect andmust listen. Both. Of course.

Pure thinking, such as mathematics or philosophy, isnot, however, to be disdained, not at all. Euler’s equa-tion, eπi + 1 = 0, really is quite remarkable, linking “thefive most important constants in the whole of analysis”(as Philip Davis and Reuben Hersh note), and would bea remarkable cultural achievement even if it had noworldly use. But certainly the equation is not a resultof looking at the world. So it is not science; it is a kind

And if you like engineers you will like many econo-mists. Engineers are attractive people, hard working(you have to be hard working to absorb all that engi-neering math), earnest and practical, bent always onSolving the Problem. True, they are often simple-minded. But simplicity gets the job done. Lots ofeconomists are engineering types.

Or lawyer types. Like lawyers the economists are goodarguers, which is good when you need a good argu-ment (“How do you want it to come out?”).Economists can debate each other and yet not losetheir tempers and not make irrelevant appeals to rank.Economists like lawyers are clear-minded, profession-ally. They are used to getting to the point and stayingthere. The humor of economists, unhappily, is oftencynical, as it is also among lawyers, seldom generous,but that’s true in many fields of the intellect.

But, above all, economics is about important matters.It would be remarkable if the economics-since-Marxthat most non-economists would rather not read hadnothing worthwhile in it. After all, thousands of appar-ently intelligent (they certainly think so) economistshave labored away at it now for a century and a half.

I beseech you, dear reader, think it possible that econo-mists, even Chicago-School economists, evenSamuelsonian economists, have some important thingsto say about the economy.

3736

car that passed. For many hours he kept it up, thrilledto be at last a real observer of society. But of coursewhen he got home and looked at the results it occurredto him that the data were meaningless. They werebrute facts unshaped by any meaningful human ques-tion, or emotion, or interest. One wishes every scholarlearned this at ten years old.

So pure mathematics, pure philosophy, the pure writingof pure fictions, the pure painting of pictures, the purecomposing of sonatas are all, when done well or at leastinterestingly, admirable activities. I have to keep saying“pure” because of course it is entirely possible—indeedcommonplace—for novelists, say, to take a scientificview of their subjects (Balzac, Zola, and Sinclair Lewisamong many others are well known for their self-conscious practice of a scientific literature; Romansatire is another case; or Golden Age Dutch painting).Likewise scientists use elements of pure narration (inevolutionary biology and economic history) orelements of pure mathematics (in physics andeconomics) to make scientific arguments. I do notwant to get entangled in the apparently hopeless task ofsolving what is known as the Demarcation Problem,discerning a line between science and other activities.It is doubtful such a line exists. The efforts of manyintelligent philosophers of science appear to havegotten exactly nowhere in solving it. I am merelysuggesting that a science like many other human practicessuch as knitting or making a friend should be about theworld, which means it should attend to the world. Andit should also be something other than miscellaneous

of abstract art. Mathematicians are proud of theuselessness of most of what they do, as well they mightbe: Mozart is “useless,” too; to what would you “apply”the Piano Sonata in A? I have a brilliant and learnedfriend who is an intellectual historian of note. He andI were walking to lunch in Iowa City one day and I saidoffhandedly, assuming he would of course know this,that mathematics was one of the great achievements ofWestern culture. He was so astonished by the claimthat he stopped short and argued with me there on thesidewalk by the Old Capitol Mall: “Surely math is likeplumbing: useful, but hardly in touch with deeperthings; hardly a cultural achievement!” I tried topersuade him that he felt this way only because he hadno acquaintance with mathematics, but I don’t think Isucceeded.

Nor is pure, untheorized observation to be disdained.There is something in narration, for example, that isuntheorizable (though it is surprising to non-humanistshow much of it can and has recently been theorized byliterary critics). At some level a story is just a story, andartful choice of detail within the story is sheer observa-tion—not brute observation, which is a hopeless ambi-tion to record everything, but sheer. I have anotherbrilliant and learned friend, an economist, who tells thestory of how as a boy in Amsterdam he decided one dayto embark in all seriousness on Social Observation. Hewas about ten years old when this ambition overcamehim, so he equipped himself with a notebook and a penand went to a big street and started to, well, observe.He decided to note down the license number of every

3938

To which I say: Bosh. She and her colleagues, when theyare being most highbrow and Science-proud, don’t really doeither theorizing or observing. Economics in its mostprestigious and academically published versionsengages in two activities, qualitative theorems and statis-tical significance, which look like theorizing andobserving, and have (apparently) the same tough mathand tough statistics that actual theorizing and actualobserving would have. But neither of them is what itclaims to be. Qualitative theorems are not theorizing ina sense that would have to do with a double-virtuedinquiry into the world. In the same sense, statisticalsignificance is not observing.

This is the double-formed and secret sin, and this themoment:

EveIntent now wholly on her taste, naught elseRegarded, such delight till then, as seemed,

In fruit she never tasted, whether trueOr fancied so, through expectation high

Of knowledge, nor was godhead from her thought.

It is not difficult to explain to outsiders what is sodramatically, insanely, sinfully wrong with the twoleading methods in high-level economics, qualitativetheorems and statistical significance. It is very difficultto explain it to insiders, because the insiders cannotbelieve that methods in which they have been elabo-rately trained and which are used by the people theyadmire most are simply unscientific nonsense, having

facts, such as the classification of animals in the ChineseCelestial Emporium of Benevolent Knowledge noted byBorges: (a.) those that belong to the Emperor, (b.)embalmed ones, (c.) those that are trained, (d.) sucklingpigs, (e.) mermaids, and so forth, down to (n.) thosethat resemble flies from a distance. Not brute facts.And not mere theory.

So I am not dragging economics over to some implau-sible definition of Science and then convicting it of notcorresponding to the definition. Such a move iscommon in economic methodology—for example insome of the less persuasive writings of the very persua-sive economist Marc Blaug. I am merely saying thateconomists want to be involved in an intelligent inquiryinto the world. If so, the field as a whole must theorizeand observe. Both. This is not controversial.

An economist at a leading graduate program listening tome will now burst out with: “Great! I entirely agree:theorize and observe, though of course as you admit wecan specialize in one or the other as long as the wholefield does both. And that, Deirdre, is exactly what wealready do, on a massive scale. And we do it very well, ifI don’t say so myself. We do very sophisticated mathe-matical theorizing, such as in the Mas-Collel,Whinston, and Green textbook (1995), and then we testthe theory in the world using very tricky econometrics,such as Jeffrey M. Wooldridge, Econometric Analysis ofCross Section and Panel Data (2001). You can see theresults in any journal of economics. Some of it is puretheory, some econometrics. Theorize and observe.”

4140

Okay, now imagine an alternative set of assumptions(like the ones used earlier to “disprove” the Free TradeTheorem), A'. Look at that last item closely. If you’regoing to venture into the wonderful world of this reallytough, macho math we economists deal in daily you aregoing to have to train yourself to look closely atsymbols: notice that the alternative assumption has alittle mark just after it, not in math called a “singlequotation mark” but a “prime” (it’s just a notation todistinguish one set of things—in this case assump-tions—from another; it has nothing to do with primenumbers). A' is read “A prime.” Naturally, if youchange assumptions (introducing households who donot operate on P-Only motivations, say; or [I speaknow to insiders] making information a little asym-metric; or [ditto] introduce any Second Best, such asmonopoly or taxation; or [ditto] nonconvexities inproduction) in general the conclusion is going tochange.

Natch. There’s nothing deep or surprising about this:changing your assumptions changes your conclusions.Call the new conclusion C' (a test of whether you’repaying attention, class: How is it read? Answer: “Csingle prime”). So we have the old A implies C andthe fresh, publishable novelty, A' implies C'. But, asthe mathematicians say, we can add another prime andproceed as before, introducing some other plausiblepossibility for the assumptions, A'' (read it “A doubleprime”), which implies its own C''. And so forth: A'''implies C'''. And on and on and on and on, until theeconomists get tired and go home.

literally nothing to do with whatever actual scientificcontribution (and I repeat, it is considerable) thateconomics makes to the understanding of society. Sothey simply can’t grasp arguments that are plain topeople not socialized in economics. (Bibliographicalnote to the insiders and the more adventuresome of theoutsiders: Chapters 10-13 in Knowledge and Persuasionin Economics [1994] and Chapters 7 and 8 in TheRhetoric of Economics [2nd ed. 1998]).

Hear, oh outsiders. I’ve told you how popular qualita-tive, Why Whether reasoning is in economics. It takesthis form: A implies C. Got it? Simple, huh? Thecrucial point is that the A and the C are indeed qualita-tive. They are not of the form “A is ‘4.8798’.” Theyare of the qualitative form, “A is ‘everyone is motivatedby P-Only considerations’,” say, which implies “freetrade is neat.” No numbers. You realize your loverwill be annoyed by the neglected birthday to somedegree, but we’re not talking about magnitudes.Why/Whether. Not How Much. The economic“theorists” focus on what mathematicians call “exis-tence theorems.” With such and such general (or notso general, but anyway non-quantitative) assumptions Athere exists a state of the imagined world C. A typicalstatement in economic “theory” is, “if information issymmetric, an equilibrium of the game exists” or, “ifpeople are rational in their expectations in thefollowing sense, buzz, buzz, buzz, then there exists anequilibrium of the economy in which governmentpolicy is useless.”

4342

have anything to do even with playing real chess (sincethe situations are often ones that could not arise in areal game). And chess itself has nothing to do withliving, except for its no doubt wonderful purity asthought, á la Mozart.

What kind of theory would actually contribute to adouble-virtued inquiry into the world? Obviously, itwould be the kind of theory for which actual numberscan conceivably be assigned. If Force equals Masstimes Acceleration then you have a potentially quantita-tive insight into the flight of cannon balls, say. But thequalitative theorems (explicitly advocated inSamuelson’s great work of 1947, and thenceforthproliferating endlessly in the professional journals ofacademic economics) don’t have any place for actualnumbers. So the “results” keep flip-flopping,endlessly, pointlessly.

The history of economic “theory” since 1947 (and, as Isaid, in non-mathematical form since 1747, too) isreplete with examples. Samuelson himself famouslyshowed in the 1940s that “factor prices” (such aswages) are “equalized” by trade in steel and wheat andso forth—as a qualitative theorem, under such andsuch assumptions, A. It could be an argument againstfree trade. But shortly afterwards it was shown (bySamuelson himself, among others) that if you makealternative assumptions, A', you get very differentconclusions. And so it went, and goes, with the limitachieved only in boredom, all over economics. Makethus-and-such assumptions, A, about the following

What has been gained by all this? It is pure thinking,philosophy. It is not disciplined by any simultaneousinquiry into How Much. It’s qualitative, not quantita-tive, and not organized to allow quantities into thestory. It’s like stopping with the conclusion thatforgetting your lover’s birthday will have some badeffect on one’s relationship—you still have no ideaHow Much, whether trivial or disastrous or some-where in between. So the pure thinking is unbounded.It’s a game of imagining how your lover will reactendlessly. True, if you had good ideas about what wereplausible assumptions to make, derived from someinquiry into the actual state of the world, the situationmight be rescued for science and other inquiries intothe world, such as the inquiry into the probably quan-titative effect of missing a birthday on your lover’sfuture commitment to you. But if not—and I’mtelling you that such is the usual practice of “theoret-ical” pieces in economics, about half the items in anyself-respecting journal of economic science—it’s “just”an intellectual game.

I have expressed admiration for pure mathematics andfor Mozart’s concertos. Fine. But economics is supposedto be an inquiry into the world, not pure thinking. (If it isto be justified as pure thinking, just “fun,” it is notvery entertaining. No one would buy tickets to listento a “theory” seminar in economics. Believe me onthis one: as mathematical entertainment the stuff isreally crummy.) The A-prime/C-prime, existence-theorem, qualitative-only “work” that economists do islike chess problems. Chess problems usually do not

4544

“But wait a minute, Deirdre,” the Insider Economistbreaks in (he is getting very, very annoyed because, as Itold you, he Just Doesn’t Get It). “You admitted thatwe economists also do econometrics, that is, formaltesting of economic hypotheses using advanced statis-tical theory. You, as an economist, can hardly object tospecialization: some people do theory, some empiricalwork.”

Yes, my dear young colleague. Since I have been toyour house and noted that you have not a single workon economics before your own graduate training Isuppose you are not aware that the argument was firstmade explicit in 1957 by Tjalling Koopmans, a Dutch-American economist at Yale (Nobel 1975), who in hisThree Essays on the State of Economic Science recom-mended just such a specialization. He recommendedthat “theorists” spend their time on gathering a “cardfile” of qualitative theorems attaching a sequence ofaxioms A', A'', A''', etc. to a sequence of conclusionsC', C', C''', etc., separated from the empirical work, “forthe protection [note the word, students of free trade]of both.”

Now this would be fine if the theorems were not qual-itative. If they took the form that theorems do inphysics (better called “derivations,” since physicists arecompletely uninterested in the existence theorems thatobsess mathematicians and philosophers), good. Thenthe duller wits like Deirdre McCloskey the economichistorian could be assigned to mere observation, fillingin blanks in the theory. But there are no blanks to fill in,

game-theoretic model and you can show that a groupof unsocialized individuals will form a civil society.Make another set of assumptions, A', and they won’t.And so on and so forth. Blah, blah, blah, blah, to noscientific end.

Such stuff has taken over fields near to economics, firstpolitical science and now increasingly sociology. Atypical “theoretical” paper in the American PoliticalScience Review shows that under assumptions A thecomity of nations is broken; in the next issue someonewill show that under A' it is preserved. This is nottheory in the sense that, say, physics uses the term.Pick up a copy of the Physical Review (it comes in fourversions; pick any). Open it at random. You will findmind-breakingly difficult math, and physics that noone except a specialist in the particular tiny field canfollow. But always, on every page, you will findrepeated, persistent attempts to answer the question HowMuch. Go ahead: do it. Don’t worry; it doesn’t matterthat you can’t understand the physics. You will seethat the physicists use in nearly every paragraph arhetoric of How Much. Even the theorists as againstthe experimenters in physics spend their days trying tofigure out ways of calculating magnitudes. The give-away that something other than scientific is going onin “theoretical” economics (and, alas, political science)is that it contains not, from beginning to end of thearticle, a single attempt at a magnitude.

So: Secret Sin Number One: qualitative theorems.

4746

of computation in the 1970s a plague in economics, inpsychology, and, most alarmingly, in medical science.Consider the decades-long dispute over the prescribingof routine mammograms to screen for early forms ofbreast cancer. One school says, Start at age 40. Theother says, No, age 50. (And still another, Neverroutinely. But set that aside.) Why do they differ?The American nurses’ epidemiological study or theSwedish studies on which the empirical arguments arebased are quite large. But there’s a lot of what engi-neers call “noise” in the data, lots of things going on.So: although starting as early as age 40 does seem tohave some effect, the samples are not large enough to beconclusive. By what standard? By the standard called“statistical significance [at the 5%, 1%, 0.1%, or what-ever level].” The medical statisticians will be glad toexplain to you (for example, the over-50 school will)that “significance” in this narrow and technical sense ofthe word tells you how likely it is the result comes justfrom the noise. A “highly” significant result is one inwhich the sample is large enough to overwhelm thenoise. That is, it’s unlikely—those 5%, 1%, etc.figures, successively more stringent—you’ll be fooledinto thinking there’s an effect when in fact the effect inthe real world is zero.

So the situation is this. The over-50 school admits thatthere is some positive effect in detecting early cancersfrom starting mammograms as early as age 40; but,they say with a sneer, it’s uncertain. You’ll be takingsome chance of being fooled by chance. Nasty busi-ness. Really, something to avoid.

no How Much questions asked, in the theory thateconomists admire the most and that has taken overhalf of their waking hours.

Still, things would not be so bad, so sunk in scientificsin, if on the lower-status empirical side of academiceconomics all was well. The empiricists like me intheir dull-witted way could cobble together actualscientific hypotheses, simply ignoring the “work” of thequalitative theorists. Actual players of chess couldignore the “results” from chess problems. In effect thisis what happens. The “theories” proffered by the“theorists” are not tested. In their stead linearizedmodels that try crudely to control for this or that effectare used. An empiricist could therefore try to extractthe world’s information about the price sensitivity ofdemand for housing in Britain in the 1950s, say.

But the sin is double. The empirical economists also havebecome confused by qualitative “results.” They, too,have turned away from one of the two questions neces-sary for a serious inquiry into the world (the other isWhy), How Much. The sin sounds improbable, sinceempirical economics is drenched in numbers, but thenumbers they acquire with their most sophisticatedtools (as against their most common tools, such assimple enumeration and systems of accounting) are itturns out meaningless.

The confusion and meaninglessness arises from aparticular technique in statistical studies, called “statis-tical significance.” It has become since the cheapening

4948

the balance of cost and benefit, since there could becosts (such as deaths from intrusive tests resulting fromfalse positives) that offset the admittedly slight gainfrom starting as early as age 40. But suppose, as waslong believed, that the costs do not offset the gain.That the net gain is slight is no comfort to the (few)people who die unnecessarily at 42 or 49 on account ofMr. Medical Statistician’s gross misunderstanding ofthe proper role of statistics in scientific inquiries. Adeath is a death. The over-50 people are killingpatients. Maybe only slightly more than zero patients.But more than zero is murder. [At this insult Mr.Medical Statistician leaps up and storms out of theroom: I told you it was difficult to persuade theinsiders; I wish I had a softer rhetoric to offer whichwould bring amoral idiots like Mr. Medical Statisticianand Mr. Econometrician around gently; but as you cansee it’s just not in me.]

Or consider the aspirin-and-heart-attack studies.Researchers were testing the effects of administeringhalf an aspirin a day to men who had already suffered aheart attack. To do the experiment correctly they gaveone group the aspirin and the other a placebo. Butthey soon discovered—well short of conventional levelsof statistical significance—that the aspirin reducedreoccurrences of heart attacks by about a third. Whatdid they do? Did they go on with the study until theygot a large enough sample of dead placebo-getters tobe sure of their finding at levels of statistical signifi-cance that would make the referees of cardiology jour-nals happy? Of course not: that would have been

Huh? Are you telling me, Mr. Medical Statistician,that even though there is a life-saving effect of earlymammograms in the data on average, you are uncomfort-able about claiming it? I thought the purpose of medicalresearch was to save lives. Your comfort is not, as Iunderstand it, what we are chiefly concerned with. Youfind the data noisy. I’m sorry God arranged it that way.She should have been more considerate. But She’sdone what She’s done. Now we have to decide if thecost of the test is worth the benefit. And your datashows that a benefit is there.

Mr. Medical Statistician, with some indignation: “Noit’s not. At conventional levels of significance there isno effect.”

Deirdre, with more indignation: Nonsense. You aretrying, alas, to make a qualitative judgment of exis-tence. Compare the poor, benighted Samuelsonian“theorist.” We always in science need How Much, notWhether. The effect is empirically there, whatever thenoise is. If someone called “Help, help!” in a faintvoice, in the midst of lots of noise, so that at the 1%level of significance (the satisfactorily low probabilitythat you will be embarrassed by a false alarm) it couldbe that she’s saying “Kelp, kelp!” (which arose perhapsbecause she was in a heated argument about a wordproposed in a game of Scrabble), you wouldn’t go to herrescue?

The relevant and quantitative question about routinemammograms, which has recently been reopened, is

5150

Journal of Economic Literature, March 1996; check it outon JSTOR; we are writing a paper examining the samejournal in the 1990s; bad news: the sin has gotten moreprevalent, not less).

The problem is that a number fitted from the world’sexperiments can be important economically withoutbeing noise-free. And it can be wonderfully noise-freewithout being important.

On the one hand: It’s completely obvious, you willagree, that a “statistically insignificant” number can bevery significant for some human purpose. If you really,truly want to know how the North American FreeTrade Agreement affected the average worker in theUnited States, then it’s too bad if the data are noisy, butthat’s not the point. You really, truly want to know it.You have to go with what God has provided.

And on the other hand: It is also completely obviousthat a “statistically significant” result can be insignificantfor any human purpose. When you are trying toexplain the rise and fall of the stock market it may bethat the fit (so-called: it means how closely the data lineup) is very “tight” for some crazy variable, say skirtlengths (for a long while the correlation was actuallyquite good). But it doesn’t matter: the variable is obvi-ously crazy. Who cares how closely it fits? For a longtime in Britain the number of ham radio operatorlicenses granted annually was very highly correlatedwith the number of people certified insane. Very funny.So?

shockingly (though not unprecedently) unethical.They stopped the study, and gave everyone aspirin. (ANew Yorker cartoon around the same time made thepoint, showing a tombstone inscribed, “John Smith,Member, Placebo Group.”)

Or consider public opinion polls about who is going towin the next presidential election. These always comehedged about with warnings that the “margin of error is2% plus or minus.” So is the claim that prediction of apresidential election six months before it happens isonly 2% off? Give me a break. What is being reportedis the sampling error (and only at conventional levels ofsignificance, themselves arbitrary). An error caused, say,by the revelation two months down the road that one ofthe candidates is an active child molester is not reck-oned as part of “the error.” You can see that a shellgame is being performed here. The statement of a“probable error” of 2% is silly. A tiny part of all theerrors that can afflict a prediction of a far-off politicalevent is being elevated to the rhetorical status of TheError. “My under streetlight sampling theory is verybright, so let’s search for the keys under the streetlight,even though I lost them in the dark.” Get serious.

The point here is that such silliness utterly dominatesempirical economics. In a study of all the empiricalarticles in the American Economic Review in the 1980s itwas discovered that fully 96% of them confused statis-tical and substantive significance (look at The Rhetoric ofEconomics, 2nd ed.; or at Stephen Ziliak and DeirdreMcCloskey, “The Standard Error of Regression,”

5352

practices of economic science and a few other fields.Economics has fallen for qualitative “results” in“theory” and significant/insignificant “results” in“empirical work.” You can see the similarity betweenthe two. Both are looking for on/off findings that donot require any tiresome inquiry into How Much, howbig is big, what is an important variable, How Muchexactly is its oomph. Both are looking for machines toproduce publishable articles. In this last they havesucceeded since Samuelson spoke out loud and boldbeyond the dreams of intellectual avarice. Badscience—using qualitative theorems with no quantita-tive oomph and statistical significance also with noquantitative oomph—has driven out good.

The progress of economic science has been seriouslydamaged. You can’t believe anything that comes out ofthe Two Sins. Not a word. It is all nonsense, whichfuture generations of economists are going to have todo all over again. Most of what appears in the bestjournals of economics is unscientific rubbish. I findthis unspeakably sad. All my friends, my dear, dearfriends in economics, have been wasting their time.You can see why I am agitated about the Two Sins.They are vigorous, difficult, demanding activities, likehard chess problems. But they are worthless as science.

The physicist Richard Feynman called such activitiesCargo Cult Science. Certain New Guinea tribesmenhad prospered mightily during the Second World Warwhen the American military disgorged its cargo to fightthe Japanese. After the War the tribesmen wanted the

In short, statistical significance is neither necessary norsufficient for a result to be scientifically significant.Most of the time it is irrelevant. A researcher is simplycommitting a scientific error to use it as it is used ineconomics and the other social sciences and in medicalscience and (a strange one, this) population biology asan all-purpose way of judging whether a number is largeenough to matter. Mattering is a human matter; thenumbers figure, but after collecting them the matteringhas to be decided finally by us; mattering does notinhere in a number.

The point is just common sense. It is not subtle orcontroversial. But thousands of scientists, and amongthem almost all modern economists, are utterlyconfused about it.

Physics and chemistry, though of course highly numer-ical, hardly ever use statistical significance (check it outfor yourself: I have in the journal Science, for example).Economists and those others use it compulsively,mechanically, erroneously to provide a non-controver-sial way of deciding whether or not a number is large.You can’t do it this way. No competent statistical theo-rist has disagreed with me on this point since Neymanand Pearson in 1933. There is no mechanical procedurethat can take over the last, crucial step of an inquiry into theworld, asking How Much in human terms that matter.

My argument is not against statistics in empirical work,no more than it is against mathematics in theoreticalwork. It is against certain very particular and peculiar

5554

ical work running on statistical significance [technicalremark: sans loss functions], through his first Ph.D.student, Lawrence Klein (Nobel 1980). Two sins, onescientist.

So it is only fair to call both the sins of moderneconomics Samuelsonian. It is rather similar to thesituation in linguistics: their Great MIT Leader isNoam Chomsky. Chomsky’s mechanical approach togrammar, fiercely denying pragmatics and therefore themain finding of the humanities in the twentiethcentury, blocks progress. So too economics. Untileconomics stops believing, contrary to its own princi-ples, that an intellectual free lunch is to be gotten fromqualitative theorems and statistical significance it willbe stuck on the ground waiting at the cargo-cultairport, at any rate in its high-end activities uninter-ested in (Really) How Much. High-end theoretical andeconometric papers will be published. Careers will bemade, thank you very much. Many outstanding fellows(and no women) will get chairs at Princeton andChicago. But our understanding of the economicworld will continue to be crippled by the spreading,ramifying, hideous sins.

Woe, woe is me. Oy vey ist mir. Pity the poor econo-mists. The sins of economics come from pride informalization, the making of great machines andmonsters:

...and called me Sin, and for a signPortentous held me; but familiar grown,

prosperity to come back. So they started a “cargocult.” Out of local materials they built mock airportsand mock transport planes. They did an amazinglygood job: the cargo-cult airports really do look likeairports, the planes like planes. The only trouble is,they aren’t actually. Feynman called sciences he didn’tlike “cargo cult sciences” (he was, ill-advisedly I think,going after sociology: apparently he was not acquaintedwith the considerable amount of good, non-statistical-significance yet quantitative and empirical and theoreti-cally meaningful sociology, such as long ago that of C.Wright Mills). By “cargo cult” he meant that theylooked like science, had all that hard math and statis-tics, plenty of long words; but actual science, actualinquiry into the world, was not going on.

I am afraid that my science of economics has come tothe same point. Paul Samuelson, though a splendidman and a wonderful economist (honestly), is a symbolof the pointlessness of qualitative theorems.Samuelson, actually, is more than merely a symbol—hemade and taught and defended the Two Sins, at onetime almost single-handedly. It was a brave stance.But it had terrible outcomes. Samuelson advocated the“scientific” program of producing qualitative theorems,developing qualitative-theorem-generating-functions (Iam making an insider’s statistical joke: ha, ha; such iseconomic humor), such as “revealed preference” and“overlapping generations” models and above all themachinery of Max U. He was involved also (it turnsout somewhat surprisingly) in the early propagation ofsignificance testing, the “scientific” method of empir-

5756

I pleased, and with attractive graces wonThe most averse.

And pity, I repeat, poor old Deirdre, who appears to bedoomed to keep making these arguments, showingmore and more plainly that the two main methods ofacademic economics are nonsense, without beingbelieved.

Cassandra, you know, was the most beautiful of thedaughters of Priam, King of Troy. The god Apollo fellfor her and made her a prophetess. In exchange hewanted sexual favors, which she refused. So he cursedher, in a most malicious way. He had already given herthe power of prophecy, to know for example whatwould happen to a science that refused to ask seriouslyHow Much. His curse was to add that though shewould continue to be correct in her prophecies, no onewould believe her.

Cassandra [to Trojan economists proposing to bring thewooden horse into the city]: The horse is filledwith enemy soldiers! If you bring it into thecity, economics is lost! Please don’t!

Leading Trojan Economist: Uh, yeah, I see what youmean, Cassie. Good point. Enemy soldiers.Inside. City lost. Qualitative theorems uselessfor a science. Statistical significance without aloss function equally useless. Economicsruined. Thanks very much for your prophecy.Great contribution. Love your stuff. [Turningto colleagues] Okay, guys, let’s bring that suckerin! �

58

Prickly Paradigm Press, LLC is the North Americansuccessor of Prickly Pear Press, founded in Britain byKeith Hart and Anna Grimshaw in 1993. PricklyParadigm aims to follow their lead in publishing chal-lenging and sometimes outrageous pamphlets, not only onanthropology, but on other academic disciplines, the arts,and the contemporary world.

Prickly Paradigm is marketed and distributed byThe University of Chicago Press.

www.press.uchicago.edu

A list of current and future titles, as well as contactaddresses for all correspondence, can be found on ourwebsite.

www.prickly-paradigm.com

Executive PublisherMarshall Sahlins

PublishersPeter SahlinsRamona NaddaffBernard SahlinsSeminary Co-op Bookstore

EditorMatthew Engelkeinfo@prickly-paradigm.com

Design and layout by Bellwether Manufacturing.