Citations of Most Often Cited Economists: Do Scholarly Books Matter More than Quality Journals?

• Published date: 01 February 2014
• Author: E. Kwan Choi,Jang C. Jin
• Date: 01 February 2014
• DOI: http://doi.org/10.1111/1468-0106.12049

CITATIONS OF MOST OFTEN CITED ECONOMISTS:

DO SCHOLARLY BOOKS MATTER MORE THAN

QUALITY JOURNALS?

JANG C. JIN*George Mason University – Korea Songdo Campus

E. KWAN CHOI Iowa State University

Abstract. This paper empirically investigates the determinants of citations based on the publication

of the top 100 most often cited economists. The effects of publication age and author fame on

subsequent citations are found to be positive and signif‌icant. Citations are also signif‌icantly affected

by popular subf‌ields in economics. However, journal quality measures, such as impact factors,

download statistics and top-4 elite journals, have insignif‌icant effects on citations. In contrast, the

citation effect of scholarly books is positive and signif‌icant, and its impact is even greater than those

of journal quality measures.

1. INTRODUCTION

Economists use academic outlets such as journals or books to disseminate their

research outputs. Based on the research outputs published in academic journals,

economists have long been ranked individually, as well as collectively (e.g.

Graves et al., 1982; Dusansky and Vernon, 1998; Jin, 2009). These rankings are

based solely on journal publications that have passed a stringent peer review

process.1It is, however, widely known that the prevailing editorial refereeing

practices especially in top-tier journals act like risk averters and thus often

exclude newly developed cutting-edge research. Quality research papers are

often published in non-refereed journals and books. This trend was even more

prominent in the old days. While working as a patent clerk, Albert Einstein

(1905) published his revolutionary work on ‘special relativity’ in Annalen der

Physik without any peer review.2Kenneth Arrow published his dissertation

‘Social Choice and Individual Values’ in 1951, and his seminal work on risk

aversion was published in the book Aspects of the Theory of Risk-Bearing in

1965.

Although inf‌luential and highly cited in the literature, scholarly books are

seldom counted as part of research productivity in the economics profession.

For example, Laband and Piette (1994a) rank the quality of economics journals,

*Address for Correspondence: Department of Economics, George Mason University-Korea Songdo

Campus, Songo, Incheon, South Korea. E-mail: jininhk@gmail.com. We thank David Ahlstrom,

David Laband and Andrew Oswald for email discussions and valuable comments on the contents of

this paper. We also thank anonymous referees and the Editor of the Journal for helpful comments

and suggestions.

1Jin (2010) shows that economic growth has a positive effect on research publications.

2The f‌irst of the so-called annus mirabilis (a Latin phrase meaning ‘wonderful year’) papers had not

received any peer reviews, nor did the f‌irst paper contain any references. In those days, the burden

of proofreading was known as an opposing side against authors. In later years, Albert Einstein had

an ongoing f‌ight with Physical Reviews.

bs_bs_banner

Pacif‌ic Economic Review, 19: 1 (2014) pp. 8–24

doi: 10.1111/1468-0106.12049

© 2014 Wiley Publishing Asia Pty Ltd

based upon ‘impact factors’ that assign citation-based quality weights only to

journals. Coupe (2003) and Kalaitzidakis et al. (2003) use a similar type of

citation-based journal quality to rank top economists worldwide. Several other

studies investigate how citations are generated. Laband and Sophocleus (1985)

and Laband and Piette (1994b) estimate the determinants of citations and f‌ind,

in both studies, a positive effect of journal quality on the number of citations.

Johnson (1997) also f‌inds that journal quality, among other factors, has a

signif‌icant effect on citations. Oswald (2007) f‌inds that the best articles in the

second-tier journals are cited more often than the average articles in top-tier

journals. All these studies in the literature focus on journal quality that deter-

mines the number of subsequent citations. No such studies, however, attempt to

analyse the role of scholarly books in citation practices.

In this paper we pose a question regarding such citations. We ask whether

publications are cited because of their intrinsic merit or because of journal

reputation. This issue has already been discussed in the published literature

(Laband and Sophocleus, 1985; Laband and Piette, 1994b; Johnson, 1997;

Oswald, 2007), but we wish to ascertain the role of scholarly books in citations.

Accordingly, we estimate the citation effects of quality journals and scholarly

books, after controlling several other variables relevant to citations. If journal

quality has an independent effect on the number of citations, such quality

measures can be a good guide to predicting the publication quality of individ-

uals, or institutions, without invoking the reputation of authors, the name value

of aff‌iliated institutions, and so forth. However, if citations are drastically

affected by factors other than journal quality, the excessive use of journal

citations would mislead educational policy-makers and university administra-

tors to undervalue research outputs, especially of scholarly books. In fact, we

f‌ind that scholarly books are highly cited, and such f‌inding would be a fresh

opening up of a neglected cause of citations in the literature.

2. MODEL SPECIFICATION AND DATA

Citations are used as a measure of research quality because they show how many

times an article or a book is actually cited by others and how much it affects

other people’s research that follows. We consider citations in a two-stage

process. In the f‌irst stage, journal articles or scholarly books are read (or

downloaded) as a searching activity in consumption. Articles published in top-

tier journals are read more often because journal quality is seemingly related to

the quality of articles. It is rather diff‌icult to judge a book by its cover; book

readership depends on other factors, such as the fame of authors. In the second

stage, only a small fraction of these are actually cited as an input to producing

articles. In this case, top-tier journal articles are most likely to be cited to impress

journal editors and reviewers. Well-known authors are also frequently cited to

indicate that the author is well versed in the literature. Thus, the number of

citations depends on the fame of authors and co-authors, as well as the quality

DETERMINANTS OF CITATIONS

9

© 2014 Wiley Publishing Asia Pty Ltd