There exists an implicit pecking order among the social sciences, and it

Transcription

1 Journal of Economic erpective Volume 29, Number 1 Winter 2015 age The uperiority of Economit arion ourcade, Etienne Ollion, and Yann Algan There exit an implicit pecking order among the ocial cience, and it eem to be dominated by economic. or tarter, economit ee themelve at or near the top of the diciplinary hierarchy. In a urvey conducted in the early 2000, Colander (2005) found that 77 percent of economic graduate tudent in elite program agree with the tatement that economic i the mot cientific of the ocial cience. ome 15 year ago, Richard reeman (1999, p. 141) peculated on the origin of uch a conviction in the page of thi journal. Hi aement wa candid: []ociologit and political cientit have le powerful analytical tool and know le than we do, or o we believe. y core on the Graduate Record Examination and other criteria, our field attract tudent tronger than their, and our coure are more mathematically demanding. At firt glance, the academic labor market eem to confirm the native judgment about the higher tatu of economit. They are the only ocial cientit to have a Nobel prize, thank to a grant from the ank of weden to the Nobel foundation. Economit command ome of the highet level of compenation in American art and cience facultie according to ureau of abor tatitic data. In fact, they earn more and have better career propect than phyicit and arion ourcade i rofeor of ociology, Univerity of California, erkeley, California, and Aociate ellow at the ax lanck-cience o Center, cience o, ari, rance. Etienne Ollion i Reearch ellow at the Centre National de la Recherche cientifique, Univerity of trabourg, rance. Yann Algan i rofeor of Economic, cience o, ari, rance. The author addree are fourcade@berkeley.edu, ollion@unitra.fr, and To acce the Appendix and Data Appendix, viit doi= /jep

2 90 Journal of Economic erpective igure 1 ercentage of Doctorate Awarded to Women in elected Dicipline, ychology ociology ife cience olitical cience and public adminitration Economic hyical cience ource: U National Center for Education tatitic, Integrated ot-econdary Education Data ytem Completion urvey. mathematician (a reeman wrote); only computer cientit and engineer do better. Unlike many academic in the theoretical cience and humanitie, many prominent economit have the opportunity to obtain income from conulting fee, private invetment and partnerhip, and memberhip on corporate board. or intance, Weyl (forthcoming) provide ome uggetive evidence that 40 percent of the income of economic author in the field of finance and indutrial organization come from conulting activitie, either with buine (finance) or government (IO). In 2010, the movie documentary Inide Job expoed the lucrative and poibly complacent relation between ome of the field mot ditinguihed member and the financial nebulae around Wall treet. Thi much better financial poition of economit, particularly in top univeritie, combined with the dicipline emphai on matering quantitative reaoning (widely interpreted a a ign of higher intellectual capabilitie) certainly tand behind the often dimiive attitude of economit toward the other, le-formal ocial cience. ut there are other reaon for the ditant relation among ocial cientit. irt, the field differ in their ocial compoition. elf-election into variou dicipline i heavily patterned by ocial attribute. or intance, economic, like phyic or philoophy but in harp contrat to ociology, i a very male-dominated field (ee igure 1). Thu, cro-diciplinary relation are inevitably permeated by broader pattern of gender difference, tratification, and inequality. And while we do not have good comparative data on the ocial origin of ocial cientit in the United tate (but ee ourdieu 1984 and ebaron 2000

3 arion ourcade, Etienne Ollion, and Yann Algan 91 on rance), we may poit that diparitie in the preent material condition of the different field generate important diparitie in lifetyle and worldview a well a relational train between them. econd, the ocial cience have experienced fat demographic growth ince World War II, which ha produced internal differentiation and hyper-pecialization. (Abbott 2001; rank and Gabler 2006). Thi proce ha obcured the common heritage the fact that hitory and moral philoophy gave birth to political economy back in the nineteenth century (ee Hakell 1977; Ro 1991, for a U-focued dicuion), while American ociology aroe partly from within economic in the early part of the twentieth century (Young 2009). In thi eay, we explore the hifting relationhip between economic and the other ocial cience in four pecific dimenion. irt, we document the relative inularity of economic and it dominant poition within the network of the ocial cience in the United tate. Though all dicipline are in ome way inular a claic conequence of the heightening of the diviion of academic labor ( Jacob 2013) thi trait peculiarly characterize economic. econd, we document the pronounced hierarchy that exit within the dicipline, epecially in comparion with other ocial cience. The authority exerted by the field mot powerful player, which foter both intellectual coheivene and the active management of the dicipline internal affair, ha few equivalent elewhere. Third, we look at the changing network of affiliation of economic over the pot-world War II period, howing in particular how tranformation within higher education (mot prominently the rie of buine chool) and the economy have contributed to a reorientation of economic toward buine ubject and epecially finance. inally, we provide a few inight into the material ituation, worldview, and ocial influence of economit, which alo et them apart from their academic peer. Taken together, thee trait help to define and account for the intellectual elf-confidence of economit and in turn for their aertive claim on matter of public policy. When we refer to the uperiority of economit, our double entendre ha both a decriptive and an explanatory purpoe. Economic occupie a unique poition among academic dicipline. It i characterized by far-reaching cientific claim linked to the ue of formal method; the tight management of the dicipline from the top down; high market demand for ervice, particularly from powerful and wealthy partie; and high compenation. Thi poition of ocial uperiority alo breed elf-confidence, allowing the dicipline to retain it relative epitemological inularity over time and fueling a natural inclination toward a ene of entitlement. While the imperialitic expanion of economic into apect of ocial cience that were traditionally outide the economic canon ha purred ome engagement with noneconomic cholarhip, the pattern of exchange remain deeply aymmetrical, cauing reentment and hotility in return. And while economit unique poition give them unuual power to accomplih change in the world, it alo expoe them more to conflict of interet, critique, and mockery when thing go wrong.

4 92 Journal of Economic erpective Inularity The intellectual trajectorie of the ocial cience dicipline have diverged importantly during the pot-world War II period. Economic ha changed ince it continental youth. It left hitory behind and turned to the paradigmatic field in the natural cience, uch a phyic, for emulation (irowki 1989). Unlike their more literary forerunner, modern-day economit attribute their intellectual tanding and autonomy to their reliance on preciely pecified and parimoniou model and meaure. They ee the field high technical cot of entry and it member endeavor to capture complex ocial procee through equation or clear-cut cauality a evidence of the dicipline uperior cientific commitment, vindicating the ditance from and the lack of engagement with the more dicurive ocial cience. In a prominent example, azear (2000, pp ) write: The acenion of economic reult from the fact that our dicipline ha a rigorou language that allow complicated concept to be written in relatively imple, abtract term. The language permit economit to trip away complexity. Complexity may add to the richne of decription, but it alo prevent the analyt from eeing what i eential. An eminent profeor echoed thi view when he decribed, thi time critically, the narrow epitemological demand of hi dicipline (interviewed by ourcade 2009, p. 91): You are only uppoed to follow certain rule. If you don t follow certain rule, you are not an economit. o that mean you hould derive the way people behave from trict maximization theory.... The oppoite [to being axiomatic] would be arguing by example. You re not allowed to do that.... There i a word for it. eople ay that anecdotal. That the end of you if people have aid you re anecdotal... [T]he modern thing [people ay] i: it not identified. God, when your cauality i not identified, that the end of you. or much of the pot-world War II period, flexing one mathematical and tatitical mucle and tripping down one argument to a formal and parimoniou et of equation wa indeed the main path to etablihing cientific purity in economic. With the empirical revolution in the 1990 and 2000, thi function ha hifted toward a hard-noed approach to cauality focued on reearch deign and inference and often extolling the virtue of randomly controlled trial (for example, Angrit and ichke 2010). Although thi move ha not ecaped criticim (for example, ee eamer 2010 and im 2010 in thi journal), it repreent a ignificant departure from the now diparaged over-theoretical orientation of the 1970 and The hift toward applied microeconomic, while very real, ha not dramatically broadened the network of interdiciplinary connection however. To be ure, economit have tarted to conider topic that are more traditionally aociated with ociology, political cience, and pychology from political intitution to family tructure, neighborhood effect, peer effect, or (a of late) ocial mobility. Yet cro-diciplinary citation pattern continue to offer evidence of the field relative inularity. Of coure, one of the mot remarkable fact about U ocial cience (continental Europe tend to be more ecumenical) i the extent to which all it contituent dicipline work in relative iolation from each other: economic,

5 The uperiority of Economit 93 ociology, political cience, and pychology all have high percentage of intradiciplinary citation. ut even o, economic tand out markedly, with 81 percent of within-field citation in 1997 againt 52 percent for ociology, 53 percent for anthropology, and 59 percent for political cience ( Jacob 2013, p. 82, who ue the N cience and Engineering Indicator 2000, online appendix 6 54, baed on a ample of the mot cited journal in each field). There are everal reaon for the inularity of economic, mot importantly the different epitemological culture of the variou ocial cience dicipline and the power inequalitie between them. irt, the theory of action that come with economit analytical tyle i hardly compatible with the baic premie of much of the human cience, namely that ocial procee hape individual preference (rather than the other way around). In economic, by contrat, de gutibu non et diputandum (tigler and ecker 1977) 1 : preference are uually aumed to be fixed (aron and Hannan 1994, p. 1116). 2 econd, the qualitative method that underpin the work of many interpretive ocial cientit often do not quare well with economit formal apiration, with their view on cauality, or with their predilection for methodological and theoretical preciion over real-world accuracy. Third, even when the ubtantive terrain overlap, the explicit or implicit pecking order between the dicipline often tand in the way of a deirable form of intellectual engagement. Examining the tructure of interdiciplinary citation in detail reveal harp difference acro dicipline. urveying academic journal from 1995 to 1997, ieter and aumgartner (2002) found harply aymmetric flow between economic and the other ocial cience. Our analyi of citation in flaghip journal for economic, ociology, and political cience over the period from 2000 to 2009 confirm thi pattern. A hown in Table 1, article in the American olitical cience Review cite the top 25 economic journal more than five time a often a the article in the American Economic Review cite the top 25 political cience journal. The aymmetry i even tarker with regard to the American ociological Review. While only 2.3 percent of the ociologit citation go to their economic colleague (often in a critical fahion, arguably), jut 0.3 percent of economit citation go to ociologit (again only taking into account the top 25 journal in each dicipline). Citation data are, of coure, likely to be biaed downward becaue ociology and political cience tend to cat their citation network more broadly overall and becaue of the role of book (which we do not account for) in thoe field. Even o, it i worth pondering thee aymmetrical pattern, epecially ince the dicrepancy i o large and other ource of evidence all point in the ame direction. A targeted comparion of citation to important figure in ociology and economic who deliberately engaged the other dicipline how thi well. rench ociologit ierre ourdieu, the top-cited name in U ociology today, received a ingle mention in the AER during the In matter of tate, there can be no dipute. 2 In the lat year however, a few economit have taken a more active interet in the formation of preference. or example, ee owle (1998) and ehr and Hoff (2011).

6 94 Journal of Economic erpective Table 1 Citation from the laghip Journal to Article ublihed in the 25 Top Journal in Each Dicipline, (a a percentage of total citation in each journal) Cited journal (% of all reference) Citing journal Top 25 economic journal Top 25 political cience journal Top 25 ociology journal Total number of paper/citation from thi journal American Economic Review 40.3% 0.8% 0.3% 907/ 29,958 American olitical cience Review 4.1% 17.5% 1.0% 353/ 19,936 American ociological Review 2.3% 2.0% 22% 399/ 23,993 ource: Compiled by the author from the electronic Intitute for cientific Information Web of ocial cience. The high number of paper and cite in the AER i due to the aper and roceeding. We alo looked at thi data without the aper and roceeding. The pattern are not ignificantly different. (againt 60 time in the AR), while Gary ecker wa reaping 41 citation in the AR (106 in AER). During the ame period ax Weber and ark Granovetter received four mention each in the AER, but Jame Heckman wa cited 25 time in the AR by ociologit, and Oliver Williamon, rom the vantage point of ociologit, geographer, hitorian, political cientit, or even pychologit, economit often reemble colonit ettling on their land an image reinforced by ome economit proud claim of economic imperialim (azear 2000). ured by the propect of a productive crop, economit are wift to probe the new ground. They may ak for guidance upon arrival, even partner-up with the local (with whom they now often hare the ame data). ut they are unlikely to learn much from them, a they often prefer to deploy their own technique. 4 And in ome cae, the purpoe ha been imply to et the other dicipline traight (Nik-Khah and van Horn 2012). Under the influence, notably, of Chicago price theory, the dominant economic paradigm ha uccefully conquered a egment of political cience, law, accounting, and (for a while) ociology under the label of rational choice theory thu explaining, in part, the directionality of the citation pattern oberved above. Opinion urvey further confirm thi analyi. Table 2 ugget that economit have in general le regard for interdiciplinarity than their ocial cientific and 3 The data come from ongoing reearch on ocial cience. or preliminary reult, ee Ollion and Abbott (forthcoming). 4 Though economit ometime alo repurpoe the technique of other, a illutrated by the borrowing of network analyi from ociology.

7 arion ourcade, Etienne Ollion, and Yann Algan 95 Table 2 Agreement or Diagreement with the ropoition: In general, interdiciplinary knowledge i better than knowledge obtained by a ingle dicipline. American univerity profeor in % Agree/ trongly agree % Diagree/ trongly diagree % No anwer/ Don t know Economic ociology olitical cience ychology inance Hitory ource: rom Gro and immon urvey about the politic of the American profeoriate. The urvey wa conducted in The author ampled 100 individual in each field. Return rate are low (though not unuually low for thi kind of urvey) and varied importantly acro dicipline (economit: 44%; ociologit: 55%; political cientit: 54%; pychologit 49%; finance profeor: 37%; hitorian: 54%). We are grateful to Neil Gro for running the cro-tabulation on thi urvey for u here and elewhere in the paper. ee Gro and immon (2007) for detail about the urvey and Gro (2013) for a broader analyi. even buine chool brethren. Economit are the only one in thi group among whom a (ubtantial) majority diagree or trongly diagree with the propoition that in general, interdiciplinary knowledge i better than knowledge obtained from a ingle dicipline. uch reult are conitent with the notion that economit, with their ditinctive confidence in the uperiority of their own dicipline, are le likely to feel the need to rely on other dicipline or even to acknowledge their exitence. A ociologit know well, thi dynamic i characteritic of unequal ituation: thoe in a central poition within a field fail to notice peripheral actor and are alo largely unaware of the principle that underpin their own domination (ourdieu 1984). Intead they tend to rationalize power and inequality a a jut product of merit, jutified by effort or talent. A good example of thi kind of rationalization would be citing higher average core on the Graduate Record Exam for graduate tudent in economic, or the higher impact factor of economic journal. ociologit, however, might point out that uch difference between field are trongly tructured by ocial determinant uch a cla, gender, and race. eanwhile, peripheral actor compulively orient themelve toward dominant one, whether poitively or negatively. 5 5 A another example of thi general phenomenon, ourcade (2006) note that non-u-baed cholar are much more likely to define their identitie around the recognition they receive (or fail to receive) from American academic intitution than the other way around.

8 96 Journal of Economic erpective Hierarchy Within The intellectual tructure within the dicipline of economic i often evoked to explain thee aymmetric relation: becaue economit have managed to preerve a more unitary diciplinary core than other ocial cience field, other field will find it eaier to refer to economic, if only to etablih a counterargument, than the other way around. In other word, the argument of a unitary dicipline are clearly identifiable from the outide, while thoe of a fractiou dicipline are more uncertain. Table 1 howed that citation in the American Economic Review are both le interdiciplinary and more concentrated than citation in the political cience and ociology flaghip journal. Thi ugget that economic more than the other field look both inward and toward the top of it internal hierarchy. Thi pattern may be interpreted in two way: there i more conenu in economic than in ociology or political cience; and there i more control. Of coure thee two interpretation are not mutually excluive: there might be more conenu becaue there i more control (for intance if a conitent view of what contitute quality reearch i promoted by thoe who control the top journal); converely, control might be more effective and enforceable becaue there i more conenu. There i ubtantial evidence that notwithtanding deep political difference amongt themelve, economit are more likely to think in a trongly integrated and unified framework than other ocial cientit. or intance, economit agree widely on the core et of principle and tool that tructure hd training. They alo rely on textbook much more than the other ocial cience do, including at the graduate level and graduate textbook tend to be written by faculty from elite department. In a urvey conducted in 1990, graduate education wa found to be amazingly imilar acro economic hd program (Hanen 1991, p. 1085). In the interdiciplinary fellowhip attribution panel tudied by amont (2009), economit had more homogeneou tandard of evaluation within, greater confidence in their judgment about reearch excellence even in other field, and a higher likelihood to tick together a a group than panelit from other dicipline. 6 Only hitorian were imilar to economit in the conitency and coheivene of their judgment about good hitorical craftmanhip, but even they were more divided internally along political line, a well a more open to conidering a variety of criteria when judging other dicipline. Judgment about the cholarly merit of propoal were more dipered and le conenual in the humanitie and other ocial cientific field, making it harder to identify important work both within and without. On the control ide, economit manage their field tightly. cholar have long noted that top department in economic exert a remarkably trong influence over the dicipline internal labor market (Cole 1983; Whitley 1984). The mot convincing 6 tudying how maintream economit etablihed their poition within the interdiciplinary chool for Advanced tudie in the ocial cience (EHE) in ari, Godechot (2011) find a imilar pattern of trong coheion within and aymmetric relation and excluion without.

9 The uperiority of Economit 97 empirical tudy on thi point come from the comparion by Han (2003) of the hiring proce in even dicipline (their tribal regime ): two from the humanitie hitory and Englih; four from the ocial cience economic, political cience, pychology, and ociology; and mathematic. Uing ingua ranca annual compilation in Job Track: Who Got Hired Where ( ), Han found, unurpriingly, that all of the dicipline follow a pretige principle : hire are trongly dependent on the pretige of department a reported by ource uch a the National Reearch Council and U New and World Report. The flow of tudent between department provide unequivocal evidence: they how that univeritie only hire from intitution that are like-ranked or higher-ranked. Academia hence reemble the kinhip ytem once decribed by Claude évi-trau ([1949]1969), in which ome alliance (between tudent and department) are preferred while other, being taboo, imply can not exit. Thi correlation between pretige and placement, however, i tronget in economic. There, the ditinction between cluter are more clearcut than in any other dicipline. Economic department at the very top of the pecking order exchange tudent amongt themelve in higher proportion than in other field, including mathematic. Three concluion emerge. irt, hierarchy i much more clearly defined in economic. econd, the field of economic i horizontally more integrated, with trong norm of reciprocity and coheion in recruitment procee. Third, thee norm utain a high tability of interdepartmental pretige hierarchie over time. y contrat, pychology and ociology are the mot decentralized, leat coheive field and have the leat table pretige ranking. Getting a Job Not imply the outcome, but alo the conduct of the annual junior job market confirm thee difference acro the ocial cience. In economic, the proce i very organized, with mot department collectively deciding on the rank ordering of their own tudent applying for poition. Thi procedure, which i uncommon in many academic field, i poible only in the context of economit trong internal agreement on quality criteria and becaue of the field belief that earch and placement procee can be more efficient that way, without altering outcome. Once a department own tudent have been ranked, market intermediarie ( placement officer ) are delegated with the tak of helping to make matche, by proactively elling the product on offer (o to peak) to potential buyer at the other end. inally, a ritualized evaluation proce progreively filter the vetted candidate, tarting with interview at the annual meeting of the Allied ocial cience Aociation held in early January. or the apiring hd graduate, the real action at the AA conference take place in the hotel uite where the hiring partie other academic department, but alo government agencie, international intitution, and private ector firm interview job candidate for everal day on end. eanwhile, in the public meeting room, the more-etablihed cholar preent their paper to their peer. The ociology junior labor market tand in harp contrat to thi careful orchetration of the circulation of tudent. To job applicant and faculty in

10 98 Journal of Economic erpective ociology, the very notion of a collectively managed proce of matching tudent to job poition would be both unworkable in practice and objectionable in principle. To be ure, ocial network play a role and informal contact ometime precede on-ite fly out, but they rarely take the form of a formal interview by a full committee, a they do in economic. Hierarchie between ociology department are alo more uncertain. A vertical tructure doe exit ociologit, too, have market tar and keep a cloe eye on commonly referenced departmental ranking. ut one would be hard-preed to define the principle that underpin the pecking order in ociology. Devoid of conenual criteria for generating a putative hierarchy, and perhap alo le truting of their colleague judgment, ociologit mut keep the proce more open in order to build up conenu from below, incluively. In economic, conenu i much tronger from the tart; information about candidate i deemed homogeneou and therefore inherently reliable. A a reult, the range of poible option i more tightly defined and determined much earlier. Getting ublihed The economic publication market i alo comparatively more concentrated than in other ocial cience dicipline in the ene that the mot-cited journal exhibit a heavier concentration of paper coming from elite department in economic than in ociology. Thi i true both in term of the department where author work and the department from which thoe author graduated. or intance, according to our calculation, the top five ociology department account for 22.3 percent of all author publihed in the American Journal of ociology, but the top five economic department account for 28.7 percent of all author in the Journal of olitical Economy ( JE) and 37.5 percent in the Quarterly Journal of Economic (QJE). The contrat i even tarker when one turn to the intitution from which the author got their hd, with the top five ociology department now totaling 35.4 percent in the American Journal of ociology, but 45.4 percent in the Journal of olitical Economy and a ky-high 57.6 percent in the Quarterly Journal of Economic. An economit might tend to regard thi concentration a evidence that, acro economic department, intellectual trength i more concentrated in the top department than i the cae acro ociology department. Other might highlight alternative metric that are alo ued for evaluation (book may be more important in ome dicipline) and the exitence of multiple criteria of worth, which are only imperfectly reflected in the hierarchy of cholarly journal. 7 Economit, by contrat, tend to ee intitutionalized hierarchie a emergent, truthful indicator of ome underlying worth, and conequently are obeed with them. or intance, in no other ocial cience can one find the extraordinary volume of data and reearch about ranking (of journal, department, and individual) that economit 7 On the role of book in academic career for ociologit, ee Clemen, owell, cilwaine, and Okamoto (1995). While the data ued in thi tudy are now 20 year old, there i no evidence that the two-pronged ituation ha changed much.

11 arion ourcade, Etienne Ollion, and Yann Algan 99 produce not to mention the centralization of economic reearch in ReEc (an international reearch archive) and the continued exitence of a ubtantial, if marginalized, ubfield focued on the hitory of economic. Thi intene awarene of hierarchie in economic breed a fierce competition for individual tatu, which may explain ome of the mot unettling apect of the field operating procedure. One notable fact i that everal leading economic journal edited at particular univeritie have a demontrable preference for in-houe author, while the American Economic Review i much more balanced in it allocation of journal pace. ooking at home bia figure ince the 1950, Coupé (2004, p. 27) find a conitent pattern of over-repreentation of in-houe author over time. etween 1990 and 2000 for intance, the Harvard-baed Quarterly Journal of Economic aigned 13.4% of it pace to it own people and 10.7 percent to neighboring IT (againt 8.8 percent to the next mot prominent department, Chicago). Converely, 9.4 percent of the page of the Chicago-baed Journal of olitical Economy went to Chicago-affiliated cholar. Thi wa equivalent to the hare of Harvard and IT combined (4.5 and 5.1 percent, repectively). Wu (2007) how that thee biae actually increaed between 2000 and Our data ( ) confirm thi domination of Cambridge, aachuett, over the Quarterly Journal of Economic and (to a leer extent) Chicago over the Journal of olitical Economy. The upremacy of Cambridge i even more triking when one look at where the author obtained their hd. In , the proportion of Harvard graduate publihing in the QJE wa 20.5 percent, jut edging IT graduate (16.4 percent). oth were way ahead of the third contributor, rinceton (7.4 percent). In the JE, Harvard, IT, and Chicago graduate all hover around percent of the author pool. To be ure, there are many reaon for home biae in economic journal, uch a higher level of ubmiion from faculty and graduate (or former graduate) tudent if the journal i edited in-houe; a higher likelihood of being encouraged by the editor, part of whoe job i to bring in good paper through interperonal connection (aband and iette 1994; edoff 2003); or journal philoophical tyle leading to elf-election biae in ubmiion. ut imilar procee are alo at play in other field without producing the ame dramatic effect. Thu, even if the ocial tructure of the field may explain ome of thee difference, it doe not explain them away: the tructure itelf tand at the core of the phenomenon that interet u here, which i the table upremacy of three department Chicago, Harvard, and IT over the ret of the field, boltered via control over two univerity-baed journal. A a point of comparion, uch home bia i virtually 8 Wu (2007) find that 14 percent of JE page publihed over that period went to Chicago author, and a whopping 28 percent of Q JE page went to Harvard IT author (pecifically, 15 percent for Harvard and 13 percent for IT). Our data for the period how that the Univerity of Chicago till rank firt with 10.8 percent of the total author publihed in the Journal of olitical Economy, followed by Harvard (6.1 percent) and the IT (4.1 percent). During the ame period, the Quarterly Journal of Economic publihed almot twice a many author (14.9 percent) from Harvard than from Chicago (7.0), with the IT coming third (6.2 percent).

12 100 Journal of Economic erpective nonexitent in the main ociology journal edited out of a univerity department, the American Journal of ociology, which i baed at the Univerity of Chicago. 9 Thi ugget that the pattern of home bia in top economic journal, together with the tability of ranking of top department, i not jut a coincidence of geography and author, but tem intead from a particular form of ocial organization and control. Getting Together inally, looking at profeional aociation acro ocial cientific field confirm the more coheive and hierarchical organization of economic, and the more fractiou character of it iter dicipline. A rapid comparion of the by-law of the American Economic Aociation (AEA), the American ociological Aociation (AA), and the American olitical cience Aociation (AA) how vat diparitie in the ditribution of political power acro the dicipline. Depite being 18,000 member trong, the AEA i a minimalit organization baed out of Nahville, Tenneee. Due are low, at $20 $40/year a of The by-law are hort, at 1,770 word, and procedure are centralized. There are only ix elected officer, and only one candidate typically run for preident-elect. A igure 2 how in dramatic fahion, the AEA leader are drawn diproportionately from the dicipline elite department: that i, 72 percent of the AEA nonappointed council member are from the top five department, in contrat with only 12 and 20 percent repectively for AA and AA. The preident-elect and program committee run the program for the annual meeting, which involve electing ahead of time the eion to be conducted and the paper from a ubet of the eion to be included in the aper and roceeding iue of the American Economic Review (the ay iue following the annual meeting). Thi procedure enure a flagging of topic and author deemed mot important by the organization leaderhip. Thi approach contrat with the more internally balkanized and alo more graroot nature of the American ociology Aociation and the American olitical cience Aociation. Although thee profeional aociation have fewer member than the American Economic Aociation (about 15,000 for AA and 13,000 for AA), their taff are larger. rocedure are more complex, a reflected in the length of their by-law: 4,657 word for the AA, 5,529 for AA. While the AEA i a unitary organization, community life among ociologit and political cientit revolve around ection or organized ubfield, each of which ha it own procedure, due, award, and program at the annual meeting. The AA olve the political problem of internal diviion by having conteted election at both the central and ection level while the AA ha long reorted to intitutionalized hore-trading between the dominant contituencie. In both cae, the organization leader are 9 If anything, our data ugget that there might intead be a bia againt Chicago faculty in the American Journal of ociology, who barely make it to the top 20 with a mere 1.4 percent of publihed paper. Although thi proportion rie to 6.9 percent for former Chicago ociology graduate, they are till topped by both Harvard hd (9.4 percent) and tanford hd(8 percent).

13 The uperiority of Economit 101 igure 2 Intitutional Compoition of the Executive Council of Three Diciplinary Organization by Ranking of Department of the Nonappointed Officer, Organization American Economic Aociation (AEA) American olitical cience Aociation (AA) American ociological Aociation (AA) ember of the council (%) Top 5 department Top 6 10 department Top department Top department Top department Unranked department Note: The ranking of department i baed on the U.. New and World Report 2012 ranking of bet graduate chool, by dicipline. The unranked category comprie motly department that do not have a graduate program, and a very mall number of foreign intitution. drawn primarily from nonelite intitution, a hown in igure 2. ecaue the diciplinary core i le identifiable and more conteted, member of the AA and the AA alo identify le with it: the rank-and-file i le bound to the elite and both aociation fulfill primarily a democratic purpoe of integration acro the board, an openne that i alo reflected in the tructuring of their conference program. However, the marginalization of mot of the aociation leader at the AA and the AA from the high-pretige core of the dicipline, and alo from political power, alo explain both organization frantic triving for influence, manifeted, among other thing, in their Wahington addree. To upport thi more elaborate infratructure and expenive reidence, due for both organization are among the highet in the ocial cience: $50 to $350/year for the AA; and $40 to $320 for the AA not counting ection due.

14 102 Journal of Economic erpective The Rie of inance or all the relative inularity and autonomy of economic, economit till do engage other dicipline. Our analyi of five top economic journal how that between 19 and 25 percent of citation are outide the dicipline, a fairly table pattern ince the end of World War II. ut when economic goe interdiciplinary, where doe it turn? Have the diciplinary connection of economic changed over time, and if o, what doe thi tell u about the evolution of the field? Thi framing provide u with a different road into the recent hitory of economic than much of the literature, which i often focued tightly on trend within economie: example include the tranformation of publihing pattern in economic journal (Card and DellaVigna 2013), the rie and fall of field within economic in volume (Kelly and ruetle 2011) and in relative pretige (Ellion 2010), or the downward trend in the ue of mathematic and in the publication of theoretical paper (Hamermeh 2013). Intead, we begin by analyzing the network of relation between economic and other dicipline over time. In other word, we tart from the aumption that who you cite ay omething about who you are. We find that changing pattern of external citation indeed tell u quite a lot about the inner ituation of the dicipline and the changing relative power of different contituencie. igure 3 offer a repreentation of economic extra-diciplinary reference, baed on our extenive tudy of citation in five top economic journal that were all founded well before World War II: the Quarterly Journal of Economic (founded in 1899), the Journal of olitical Economy (1899), the American Economic Review (1911), Econometrica (1933), and the Review of Economic tudie (1933). 10 The figure tell a tory that i partly familiar, partly le o. The point in the figure how the hare of outide-the-field citation in economic journal going to journal in the field of finance (), tatitic (), buine (), political cience (), mathematic (), ociology (), and law (). ecaue there i coniderable fluctuation from year to year, we how the pattern of the data a moothed curve. The figure how the dramatic rie of economic engagement with mathematic and tatitic in the pot-world War II period. The high point of thi engagement, in the mid-1970, coincide with the low point of engagement with the other ocial cience (uch a political cience and ociology), a well a with practical enterprie, uch a law and, with a light delay, buine. Notwithtanding the foundation and government effort to promote interdiciplinary venture under the behavioral cience label in the 1950, the ocial cience became clearly more etranged from one another in the Nor wa economic the only driving force in thi proce: cro-diciplinary experiment at Harvard (the Department of ocial Relation) and Carnegie-ellon failed, and all the variou field retreated into their own ditinctive form of abtraction and high theory (teinmetz 2005; Iaac 2010). 10 Citation were obtained from the Intitute for cientific Information Web of ocial cience. The line were drawn uing a moothing coefficient. ee the online Appendix available with thi paper at for detail.

15 arion ourcade, Etienne Ollion, and Yann Algan 103 igure 3 Extradiciplinary Citation in ive Top Economic Journal (to paper in field of finance, tatitic, buine, political cience, mathematic, ociology, and law) 8 % of citation inance () 4 tatitic () uine () olitical cience () athematic () ociology () aw () ource: The raw citation data were collected from the Intitute of cientific Information Web of ocial cience. Note: The point in the figure how the hare of outide-the-field citation in five economic journal going to journal in the field of inance (), tatitic (), uine (), olitical cience (), athematic (), ociology (), and aw (). The top five economic journal are the Quarterly Journal of Economic (founded in 1899), Journal of olitical Economy (1899), American Economic Review (1911), Econometrica (1933), and Review of Economic tudie (1933). We how the pattern of the data a moothed curve. The line were drawn uing a moothing coefficient. ee the online Appendix available with thi paper at for detail. The interdiciplinary ecology a it tood toward the end of the period depicted in igure 3 look very different. Citation to mathematic in the leading economic journal are practically gone and thoe to tatitic have faltered. The other ocial cience have made a modet comeback, particularly political cience (which ha had a partial converion to rational choice theory). ut the mot triking trend from igure 3 in recent decade i the continuou rie of finance a a purveyor of interdiciplinary reference for economic. In judging the magnitude of thi trend toward finance, it i important to note that our etimate of the rie of the role of finance within economic in igure 3 i very conervative. Our lit of five top economic journal doe not include any finance journal. igure 4 preent an analyi of citation among our lit of five top economic journal plu two more: the Journal of inance (founded in 1946) over time; and the ritih-baed Economic Journal (founded in 1891) a core generalit publication for economit for much of the twentieth century, on par with the JE and Q JE at the beginning of the period. elf-citation are not counted in the total of cro-citation. Reading the graph, we ee that in , the AER got 33 percent

16 104 Journal of Economic erpective igure 4 Citation among ix Economic Journal and One inance Journal (excluding elf-citation; (Quarterly Journal of Economic (QJE), Economic Journal (EJ), Journal of olitical Economy (JE), American Economic Review (AER), Econometrica (ECTRA), Review of Economic tudie (RE), and Journal of inance (JIN)) 30 % of citation received Cited AER JE ECTRA QJE RE JIN EJ ource: The raw citation data were collected from the II Web of ocial cience. Note: igure 4 preent an analyi of citation among ix economic journal plu the Journal of inance. elf-citation are not counted in the total of cro-citation. Reading the graph, we ee that in , the AER got 33 percent of the cro-citation among that et of journal, elf-citation removed. In addition to the data point, the graph how line drawn uing a moothing coefficient. of the cro-citation among that et of journal, elf-citation removed. The graph how a lot of action at the top the meteoric return of the QJE to prominence, the relative decline of Econometrica and JE but two other alient tranformation over the very long run are the contant decline of the ritih journal (RE and EJ), particularly the EJ (which diappear into near-oblivion) and the rie of the Journal of inance. Our bibliometric network data (not hown) indicate that by the 2000, the J wa mot cloely integrated with the core U-baed publication, receiving between 7 and 11 percent of all the cro-reference (excluding elf-citation) in the AER, QJE and JE. In other word, the J, which would not have been conidered an economic journal when it wa firt founded, ha become an integral part of the economic diciplinary matrix. Other finance journal have followed uit, too, a financial economic ha become the dominant approach in the field ( Jovanovic 2008) The firt iue of the Journal of inancial Economic came out in 1974, and it i now ranked a the eighth economic journal by impact factor. The Review of inancial tudie, firt publihed in 1988, rank twelfth. The Journal of inance now rank fifth by impact factor overall in economic, edging ahead of the much older Review of Economic tudie.

17 The uperiority of Economit 105 The intitutional rie of finance a an intellectual powerhoue within economic follow from the etablihment of a teaching bae in buine chool in the econd half of the twentieth century. Over that period, buine chool, which control the production of certified manager (through the A degree), have evolved from practitioner-dominated program truggling for academic legitimacy to become the larget employer of trained ocial cientit, now rivaling traditional academic department in the ize and ditinction of their facultie. A urvey from 2004 found 549 economic hd teaching in the top 20 U buine chool, a compared with 637 economic hd in the top 20 economic department (lau 2006). Thi aborption of increaingly large contingent of economic hd ha turned buine chool into formidable player within economic cience itelf a tranformation that i atteted by the remarkable tring of Nobel rize in economic cience awarded to cholar baed in buine chool ince 1990 (ourcade and Khurana 2013), including Eugene ama, Oliver Williamon, Robert Engle, ichael pence, Robert erton, yron chole, erton iller, John Haranyi, and Robert ogel. Our own analyi of paper publihed in the American Economic Review ince the 1950 reveal a rapid rie in buine chool affiliation among author, and a imultaneou and harp decline in government-baed author. The hare of author whoe primary affiliation i to a buine chool ha increaed teadily from a low 3.2 percent in the 1950 to 17.9 percent in the Converely, contribution from cholar located in government agencie have become marginal. 12 A the academic field of economic hifted toward buine chool and away from government economit faced a new et of practical, intellectual, and political entanglement: higher level of compenation, new connection and conulting opportunitie, and often different politic a well ( Jelveh, Kogut, and Naidu 2014). In the 1980, upicion of government action grew markedly within the field, and economit arguably upplied part of the intellectual rationale for the deregulatory movement in public policy and for the expanded ue of price and market mechanim in education, tranportation, healthcare, the environment, and elewhere (lyth 2002). inancial economit argued forcefully that the purpoe of corporation wa to maximize hareholder value, and provided a cientific jutification for the management practice favored by a new generation of corporate raider: leveraged buy-out, merger and acquiition, and compenating corporate executive with tock option. 13 In a recent indictment of the pervaivene of the capture of economit by buine interet, Zingale (2013) found that, when none of their author worked in a buine chool, economic article were ignificantly 12 eaure are baed on elf-declared affiliation on the article we urveyed. When author mentioned everal affiliation (a trait that ha increaed over time), we adopted the following procedure: If there wa a clear order, we opted for the firt intitution. Otherwie, and in an attempt to not artificially increae the hare of econdary affiliation, we gave priority to economic department when mentioned equally with any another intitution. ee the online Appendix available with thi paper at for detail. 13 or intance, ee ligtein and hin (2007), Jung and Dobbin (2012), ourcade and Khurana (2013), and Heilbron, Verheul, and Quak (2014).

18 106 Journal of Economic erpective le likely to be poitive on the level of executive compenation, and ignificantly more likely to be negative (p. 139). 14 A ife of Their Own Economit have ditinctive opinion, belief, and tate compared to academic in other field and to the broader American public. Evidence on thi topic i dipered and mut be pieced together from variou ource. A izeable hare come from economit themelve: the home-grown literature on the topic i abundant. The field i filled with anxiou intropection, prompted by economit feeling that they are powerful but unloved, and by robut empirical evidence that they are different. In ome claic example, arwell and Ame (1981) found that firt-year graduate tudent in economic at the Univerity of Wiconin were le likely to make contribution to a public good in a tructured laboratory game. In thi journal, rank, Gilovich, and Regan (1993) cite a range of evidence uggeting that tudying economic inhibit cooperation. The extent to which uch difference perit acro the context of different laboratory tudie, and the underlying caue of any difference that do perit, remain controverial. I it that learning economic make people more accepting of elf-intereted behavior in themelve and other? Or perhap it i that the dicipline attract more egoitic people? rey and eier (2005) look at voluntary tudent contribution to ocial fund at the Univerity of Zurich, and find that thoe who will later chooe economic a a field of tudy are le likely to contribute even before their economic tudie begin. Whatever the underlying dynamic, there i uggetive and convergent evidence that economit are either more candid about puruing their elf-interet, or imply more elfih (by dipoition or a a reult of training). Economit are likely to find themelve in a minority poition on ome of their dearet idea. apienza and Zingale (2013) argue that the more American economit agree among themelve, the more ditant they grow from average American. In general, of coure, economit favor uing market-baed olution to addre ocial iue (Whaple 2009). They upport allowing payment to be made to organ donor, but the public find the very thought ditateful. A izeable majority of economit believe that trade protectionim i economically harmful, but when aked whether buying American i good for the economy, the average American agree it i (apienza and Zingale 2013, p. 638). Economit think that a market mechanim uch a a carbon tax or a cap-and-trade ytem of pollution permit i a more cot-effective mechanim to curb climate change than regulatory tep uch a car emiion tandard, but mot of their fellow citizen beg to differ. Economit may advie government, but they often do not convince the people. 14 The ample included 150 of the mot cited downloaded RN paper prior to 2008 uing the earch key word executive compenation (excluding urvey paper).