In preparation for a PhD class, I decided to look up what the most influential papers in economics are as of late 2023.1 I wanted to learn what ideas and papers have been most influential. (This is not necessarily the same as “the best” though I do not know how one would objectively measure this.).
Most Influential Economics Papers
The Web of Science makes it easy to pull down the 200 most-cited economics papers. Just use a simple query “wc=(economics)”, query link.
The November 2023 results are here in xlsx spreadsheet form that anyone can use to make their own list. In web html form, the papers are ordered as [a] high-impact papers without Nobel prizes attached; [b] high-impact papers with Nobel prizes attached; [c] econometrics papers; [d] others.
Nobel Prize Validation
First, let’s see if the committee is even interested in impact. The following figure plots the 50 most-influential papers in economics against their age, as of 2022 (i.e., last year).
Most of the papers on the list had an author who has already won a Nobel Prize, is dead, or both. Such papers are shown with a cross in the graph. Econometric papers are plotted somewhat smaller, because citations in econometrics are hugely overrepresented relative to the number of researchers in the area. Nevertheless, and of course, a large number of authors of econometrics papers have (appropriately) also won a Nobel Prize already. The yellow dots, which are annotated with a paper abbreviation, are remaining omissions. They suggest some good candidates.
Future Nobel Prize Winners?
1. Theory of Firm and Agency: Michael Jensen (plus)
Disclaimer: I have met Michael Jensen once, but do not really know him. I have never been one of his followers or circle. I doubt he would recognize me.
The clear outlier of most-cited economist not on the list of prize winners is Michael Jensen. His 1976 Jensen-Meckling paper (jm76) laid the foundation for the modern (agency-based) theory of the firm in financial economics. His 1986 AER paper (j86) on takeovers remains similarly influential. (His Jensen-Murphy paper on executive compensation comes in at #161, too.) Moreover, Jensen has managed to put together his career away from the centers of academic power, which was much more difficult to accomplish. Jensen was not academic royalty (i.e., at a top-5 institution).
One reasonable choice would be to add Stewart Myers to this Nobel Prize. His Myers-Majluf paper (mm84) appears on #33, and also covers roughly the same ground, i.e., capital structure and issuing activity.
2. Growth and Institutions: Andrei Shleifer and Daron Acemoglu (plus coauthors?)
Disclaimer: I know both Andrei and Daron. I am friendly but not friends. I am pretty sure they would recognize me. I have never been one of their followers or circle. They are absolutely dazzling and amazing scholars.
Andrei and Daron have founded the modern empirical analysis of political economy and institutions, with a cross-country emphasis. LLSV 1998 (“Law and Finance”, llsv98) ranks at #24. A reasonably choice to add would be Robert Vishny. (The papers Andrei has written with Rob are generally better than the papers he has written without him.) Both Andrei and Daron have further published very influential papers in other areas, too. Acemoglu-Johnson is ranked about #70.
Unlike Jensen, both Andrei and Daron have always been royalty — in this case, more than deserved. In a sense, they have founded royal houses. They have made Chicago, MIT, and Harvard as good as they are.
3. Panel Econometrics: Stephen Bond (plus)
Disclaimer: I don’t know what I am talking about here. The area is too far from my expertise. I have never met Steve.
Econometrics is difficult for me to judge. Most econometric papers on the list of the top 200 economics papers are variations on unit roots in time-series. Arguably, panel methods have not received as much attention. Bond has two papers on the top-20 list: #6 (Arellano-Bond) and #11 (Blundell-Bond). (I don’t know the relative contributions of Arellano and Blundell to these papers. These two authors may be appropriately included, too.)
4. Others.
-
Fairness: Fehr and Schmidt are very impressive in not having authored their hit paper at a royal court. The paper is also amazingly recent.
-
Economic Growth: There are a large number of papers related to growth on the top 200 list. They include Mankiew, Romer, Weil; Barro, Grossman, Krueger, Aghion, Howitt, Mauro, Hall, Jones, Rajan, Zingales
-
Market Microstructure: Pete Kyle, perhaps the smartest guy on the list (and that means a lot!!), has one huge hit conceptual theory paper (#98).
-
Information and Data
Historical Association
Citation influence is a nice objectively measurable criterion, but it isn’t everything. For example, in my mind, the Nordhaus Nobel Prize is difficult to beat, yet Nordhaus’ highest-impact paper is only about about #400. Similarly, Piketty should probably eventualy win a Nobel Prize, even though his book on inequality is nowhere close to the top here. (Piketty-Saez is ranked around #500.)
Some past Nobel prizes seem to have been given more for “life-time contribution” than for “great ideas.” Others seem to have been given for perceived importance beyond cites (such as Nordhaus). Finally, yet others (especially in the distant past, such as Richard Stone) seem difficult to explain to me.
In 2021, I also spent some time looking at the top-200 papers, not just the plotted top-50 papers. As in the graph based on the top 50 or so papers, the top 200 paper could similarly be grouped into three categories:
- they are econometrics/methods;
- the authors have a Nobel prize, or
- the authors are not in the suitable age category (too young or too dead).
Past Associations of Influence with Nobel Prizes
Not all Nobel Prize winners have had highly influential papers on the top 200 list. In the following, I add the highest cited paper for each Nobel Prize winner:
Have Top-200 Paper | Have No Top-200 Paper |
---|---|
- Yes: akerlof (#16), angrist (#31), arrow (#89), banerjee (#182), becker (#37), coase (#9), diamond (#136), dybvig (#136), engle (#4), fama (#11), hansen (#40), hart (#93), hayek (#63), heckman (#3), holmstrom (#103), kahneman (#1), krugman (#71), kuznets (#76), kydland (#151), lucas (#18), markowitz (#7), merton (#67), miller (#47), modigliani (#47), nash (#84), north (#109), prescott (#151), romer (#23), samuelson (#138), scholes (#6), schultz (#163), sharpe (#33), simon (#38), sims (#64), solow (#25), spence (#46), stigler (#70), stiglitz (#62), thaler (#174), tobin (#105), vickrey (#101), williamson (#85). | - No: bernanke (#307), deaton (#275), debreu (#280), diamond (peter, #405), fogel (>#1000), friedman (#217), frisch (>#1000), goldin(>#1000)., haavelmo (>#1000), harsanyi (#517), hicks (>#1000), hurwicz (>#1000), imbens (#294), kantorovich (>#1000), klein (>#1000), koopmans (#977), kremer, (>#1000), leontief (#623), lewis (>#1000), maskin (#710), mcfadden (#272), meade (>#1000), milgrom (#200), mortensen (#576), mundell (#304), myerson (#779), myrdal (>#1000), nordhaus (#415), ohlin (>#1000), ostrom (#356), phelps (#474), pissarides (#576), roth (#957), sargent (>#1000), schelling (>#1000), selten (>#1000), sen (#986), shapley (#231), smith (>#1000), stone (>#1000), tinbergen (>#1000), tirole (#382), wilson (#496). |
More Observations
Among the top 200 papers:
-
The single-most cited work in economics is Kahneman and Tversky’s prospect theory.
-
The most common authors were Shleifer (8), Fama and Vishny (7), Merton and Jensen (5).
-
The youngest entry on the list is the Wooldrige textbook (2010, #35). With the exception of Reinhart-Rogoff (“This time is different”, 2009, #156), all young papers are econometric methods — but, of course, RR is based on a spreadsheet error. the Graham-Harvey-Rajgopal (2005, #151) survey of corporate officers is the youngest other paper (#192).
-
The oldest entry on the list is Hotelling’s competition paper (1929, #79). The next oldest paper is Coase (1937, #10). Talking about staying power!!! These papers have truly stood the test of time.
-
Most of the papers are by American or British authors.
-
Almost exactly one-third of the papers on the top-200 list are on econometric methods, such as panel and unit-root methods. The number of highly-cited papers is greatly overrepresentative relative to the number of econometricians in economics. (They provide useful tools to all of us, in the best sense!)
-
About one-third of the researchers in the broader economics profession are in finance, based on membership in the AEA vs. AFA. This is approximately propotionally reflected in the appearance of finance papers on this list. About one-sixth are asset-pricing papers, just under one-sixth are corporate finance papers.
-
About one-sixth of the paper are in macro- and growth economics.
-
About one-quarter of the papers are (primarily) economic theory. This area distribution seems reflective of the past, more so than the present.
-
Marc Melitz is young but should absolutely eventually win for the Melitz model of trade (already ranked #53 in 2023!). Amazingly, the paper states that it “adapts Hopenhayn’s (1992a) dynamic industry model to monopolistic competition in a general equilibrium setting.” Maybe this means that Hopenhayn should be a part of this.
-
Carhart’s 1997 paper is listed as about #36. However, it is not a good candidate for a Nobel prize, the same way that lots of econometrics papers are not. The paper provides a “specification” control inclusion for the momentum factor.
-
Only 3 of the top 200 papers include a female author: Bertrand-Duflo-Mullainathan (#75^2). Reinhart-Rogoff (#159). Hart-Laury (#171).
- It seems less plausible that women are so few because there was an active coordinated campaign to discriminate against women. Cites are not decided by a small clique (e.g., of men as editors or referees or department chairs). Women are of course as good as economists as men. It seems more likely to me that the dearth of women on this list reflects the historically smaller number of women in economics. Feel free to disagree.
Some of the papers/books whose validities are disputed are:
-
Reinhart-Rogoff (see Herndon-Ash-Pollin 2014; but read also the response).
-
some of the LLSV papers (see Holderness 2016a and Holderness 2016b; no response); and
Final Notes
-
Armen Alchian, Harold Demsetz and Stephen Ross absolutely should have won Nobel prizes when they were alive. How could Richard Stone have won but not Harold Demsetz?
-
Bollerslev [ranked paper #15] should have snuck in with the ARCH Nobel prizes in the past, but it’s probably too late now.
-
I also like Newey-West (#18), but it seems conceptually too derivative to White (#4) even though NW is so widely used. On the other hand, it took 7 years to publish NW after White. Perhaps NW just seems obvious in retrospect.
-
There is a whole universe of papers in ecology, evolution, and economics (e.g., Eric Charnov or Robert Costanza). I am not qualified to judge the relative contributions here, but this literature seems fascinating.
Feel free to drop me a note if I missed something important.
Stupid Dinner Table Trivia
What are the most over- and under-represented beginning letters of the last name of recipients?
Last Name | US Pop | Nobel |
---|---|---|
A | 3.75% | 3 |
B | 8.96% | 3 |
C | 6.38% | 1 |
D | 5.65% | 5 |
E | 2.03% | 1 |
F | 3.63% | 4 |
G | 5.34% | 1 |
H | 5.59% | 9 |
I | 0.76% | 1 |
J | 1.36% | 0 |
K | 5.70% | 8 |
L | 5.24% | 3 |
M | 8.28% | 12 |
N | 2.13% | 3 |
O | 1.63% | 2 |
P | 5.27% | 3 |
Q | 0.26% | 0 |
R | 4.80% | 2 |
S | 10.93% | 16 |
T | 3.88% | 4 |
U | 0.47% | 0 |
V | 2.55% | 1 |
W | 3.46% | 2 |
X | 0.02% | 0 |
Y | 0.65% | 0 |
Z | 1.29% | 0 |
Overrepresented?: S, M, H. F, A.
Underrepresented?: J, Q, U, X, Y, Z. (Well, the Chinese have only more recently entered the profession in large numbers… ;-) )
Footnotes
-
RePeC is not my favorite, because a large part of its scholar rankings is based on authors’ linkage with others. It thus measures “pedigree” as much as it measures “merit.” If anything, the profession should move away from pedigree and more towards merit. (I have benefitted from pedigree, too, though I am only “aristocracy” and not “royalty.”) Moreover, RePeC’s input data is pretty poor, especially with respect to financial economics.
Clarivate’s Web of Science and Google Scholar are both pretty decent sources of citation counts, although they have different normalizers. The rankings are comparable, the raw counts are not. The ISI highly-cited researcher lists are simply terrible, incomprehensible, and irreproducible. Whoeever is in charge of them deserves to be fired.
I deleted Nocedal-Wright (an optimization textbook), the Zurich experimentation toolbox, but left others: The computer movie simulation of Detroit growth from 1970; the econometrics textbooks (angrist-pischke, wolldridge). I removed duplicate entries for pedroni and maddala, and deleted a bookintro (or error?) by Rosewarn. ↩