Reputational capital and academic pay.

Moore, William J. ; Newman, Robert J. ; Turnbull, Geoffrey K. 等

GEOFFREY K. TURNBULL (*)

I. INTRODUCTION

Economists continue to have a Strong interest in academic labor markets, as the steady stream of earnings studies over the past several decades readily attests. The academic labor market draws the attention of economists not only because it is a topic of direct personal interest, but also because it presents an opportunity to examine general theories of labor markets in a context with which the researcher has intimate experience. Nonetheless, important questions about how the market values research, teaching, and service remain unanswered. To what extent does the market reward quality versus quantity of published research or team production in the form of co-authorship? Furthermore, how durable is reputational capital for individuals whose research productivity diminishes during their career? Ultimately, these questions relate in one way or another to the nature of the incentive structure faced by academic economists.

We provide new evidence concerning the role of research quality, co-authorship, administrative service, and teaching performance in the production of reputational capital. In addition, we explore the question of the durability of reputational capital once some of these activities cease. We assume that the quantity and quality of published research largely determines an individual's reputational capital in the external market. To help us sort out the separate effects on reputational capital of both the quality and quantity of research activity, we measure research quality by both journal quality and career citations, employing an interactive quantity-quality term.

A related question concerns the role of team production in building an individual's reputational capital. Do individuals receive differential returns to publishing single-authored versus co-authored research papers? This is an interesting question in light of the substantial increase in co-authorship over the last few decades. (1) One explanation of this trend is that the academic market has presented scholars with increasing opportunities for specialization in the production of knowledge and that team production involves the combination of complementary-skilled researchers. If team production reflects complementarities in the production of research, we should observe little difference in the returns to single- versus co-authored work. Another explanation is that co-authorship involves the combination of substitute-skilled researchers so that co-authorship presents a way for individuals to reduce their research efforts without reducing their individual measured research output. To the extent that co-authorshi p represents such substitution, the market should fully discount the returns to multiple-authored papers.

In this article we also move beyond identifying the factors that enhance an individual's reputation in the market to study the question of the durability of reputational capital. Some individuals quit publishing once they are tenured or promoted while others experience lengthy but temporary breaks in their flow of research output. We examine the potential penalties associated with gaps in research output, providing the first empirical evidence concerning the durability of reputational capital for academic economists.

In order to ensure that our analysis deals with these issues in a way that is most relevant to a large portion of the market for academic economists, we use data from several research universities, but restrict the sample frame to a single discipline: economics. Much of the previous work estimating how publications or other factors affect academic salaries typically use data from a single university, with earnings equations estimated across disciplines. Restricting the sample to a single discipline has the additional advantage of reducing potential heterogeneity problems associated with measuring productivity across disciplines. In addition, our data enable us more thoroughly to explore the empirical issues in the debate, since they permit more precise measures of the past and current productivity of individuals engaged in comparable work. Thus, it is possible to compare the relative earnings and relative productivity of individuals in fairly homogeneous jobs.

II. DATA AND VARIABLE DEFINITIONS

Data for this study are taken directly from individual vitae of full-time, tenured faculty members at nine public Ph.D. programs in economics for the 1992-93 academic year. The sample excludes individuals who currently occupy administrative positions above the level of department chairperson. According to a recent study by Scott and Mitias (1996), these graduate programs ranked from approximately thirtieth to seventy-fifth in the United States. While many of the previous studies examined earnings differentials at top-tier programs, we feel that the universities in our sample are representative of departments that constitute a large part of the market for academic economists. It would be a surprise, moreover, to discover that relative earnings of academic economists depend on productivity differentials only at top-tier programs.

The dependent variable is the log of the 1992-93 academic year 9-month salary. We do not include consulting fees, royalties or other income earned outside the university. The independent variables fall into two categories: personal demographic characteristics and productivity measures.

Demographic Characteristics

Professional EXPERIENCE is measured in years of academic employment subsequent to date of highest degree. Job tenure or SENIORITY is defined as the number of years employed at the current institution beyond the level of a visitor or non-tenure track position. SEX is a binary variable equal to unity if the individual is male. FOREIGN is a dummy variable indicating whether the individual's bachelor's degree is from a non-English speaking country. Finally, PH.D. PEDIGREE is a dummy variable that is equal to unity for an individual whose Ph.D. was earned at one of the top graduate schools. (2) Among other things, this variable may capture unobservable quality dimensions of a heterogeneous faculty. It may also capture any lingering "halo effects" associated with the individual's degree.

Productivity Measures

We construct several measures to capture faculty productivity in three dimensions: research, teaching, and administration. While the weight attached to each depends on a department's objective function. Our sample consist of departments that emphasize research. Thus we assume that that objective function of each department in our sample attaches a relatively heavier weight to the research component in its salary determination. We use a variety of measures to disentangle the quality and quantity dimensions of each individual's research record.

In academic labor markets the cumulative stock of published research represents one direct measure of an individual's reputational capital. The broadest measure of reputational capital is captured in the variable TOTAL PUBLICATIONS, defined as the sum of books, manuscripts, and journal articles. We partition journal articles into two groups to provide a measure of quality. The variable LEVEL I ARTICLES measures the number of papers published in American Economic Review, Econometrica, Economic Journal, Economica, International Economic Review, Journal of Economic Theory, Journal of Political Economy, Quarterly Journal of Economics, Review of Economic Studies, and The Review of Economics and Statistics. Although the profession offers no unanimous view on exactly which journals belong in the top-tier category, this list contains the most significant outlets for departments in our sample. (3) Also, while a couple of these journals may have slipped in prestige, many of the senior scholars in our sample were publis hing during the years in which these journals would certainly have been included in the premier group. Their present earnings reflect increments that were, in part, determined by the relative quality of those journals at that time.

The variable LEVEL II ARTICLES is the number of papers published in respected second-tier general interest journals and the top field journals. Our selection of the 45 journals in this group was guided by the impact-adjusted citations per capita rankings by Liebowitz and Palmer (1984), but also includes some newer field journals that were not included in that study (see appendix). (4)

We include CAREER CITATIONS as an additional quality measure of published research and reputational capital. Like Hamermesh et al. (1982), we argue that the frequency of references to an individual's scholarly work, at least in part, measures the influence of an individual's research on the profession. We interpret citations as evidence that the individual's work is important to the profession.

We construct CAREER CITATIONS using the individual citations reported in the SOCIAL SCIENCE CITATION INDEX (SSCI), matching citations to vitas paper-by-paper in order to assign complete citation counts to co-authors. Because we want to measure the effects of original research on the profession, we omit self-citations and citations to textbooks. To the extent that citations and quality are positively correlated, citations reflect one dimension of an individual's reputational capital and hence, tend to increase earnings. (5) The effect of citations (reputation) on salary does not require that administrators know individual faculty members' actual citation counts. Rather, the effect is more likely to occur because of the market competition for individuals with outstanding reputations.

The variable YEARS AS CHAIR is included in the model to capture the earnings differential attributable to an individual's accumulated administrative skills. We define this variable as the number of years an individual has served as a department chair, either currently or previously. (6) Whatever the reason for paying individuals a premium to serve as chair, our experience suggests that departments do not randomly select individuals for this position. Therefore, the length of an individual's service likely reflects administrative productivity since unproductive chairs lose faculty support and are replaced more quickly than their more productive counterparts.

Finally, we proxy the quality of teaching by TEACHING AWARD, a dummy variable indicating whether an individual has ever received an award for outstanding teaching. While the reward structure in research-oriented departments arguably places more weight on research output, we anticipate that high-quality teaching is valued as well. (7)

Table 1 reports summary statistics for the faculty in our sample. On average, full professors have about 21 years of experience in the profession and they have spent 16 years (approximately 75% of their careers) at their present university. Most publish regularly, with an average of 45 career publications. About 8 percent of these publications are in the top-tier journals, and 22 percent are in the second-tier journals. Six percent of the full professors have published 11 or more top-tier articles over their careers. The maximum number of top-tier publications is 20. Clearly, our sample of full professors consists of individuals publishing regularly in level II journals; only 8 percent of full professors failed to produce a level II article, while 10 percent have produced 18 or more level II articles. The average number of career citations for full professors in our sample is 147 (two individuals have over 1000 citations each). This represents about 3 citations for each publication. Typically, however, citati ons to an individual's published research tend to be concentrated on a relatively small number of papers. For example, for one of the individuals receiving over 1000 career citations, three papers account for over 32 percent of the total number of citations.

The associate professors in our sample have about 14 years of professional experience, 12 of which have been spent at their current institution. Approximately 13 percent of their total publications appear in level I journals and 25 percent in level II journals. Four percent of associate professors have five or more level I articles, and 11 percent have seven or more level II articles. Only four associate professors have failed to publish at least one level II article. Overall, 32 percent of the total publications for faculty in these departments appear in the top-tier and second-tier journals.

III. THE EMPIRICAL RESULTS

Model (1) in Table 2 presents the ordinary least-squares regression estimates of the logarithm of nine-month salary on the independent variables discussed above, excluding proxies for reputational capital, and other productivity measures. (8) This model represents a variant of the standard human capital earnings function. The estimates depict an increasing concave experience-earnings profile for academic economists. The quadratic experience terms are jointly significant ([F.sub.2, 128] = 28.58). At this stage, the point estimate for SENIORITY is negative and significant, a result that is consistent with many previous studies on the nature of earnings functions in academic labor markets. (9) The coefficients on the other demographic variables are insignificant, however.

In model (2) we introduce total publications and the quality-adjusted measures of research as the proxies for reputational capital. We also include an interaction term for citations and total article production to determine whether reputational capital is related to the way in which the stock of citations is generated. That is, does the market distinguish between scholars who generate a large number of citations from a relatively small number of papers and scholars who generate the same number of citations by writing a larger number of papers? To the extent that citations reflect the quality of one's research, we expect that individual papers receiving a large number of citations enhance a scholar's reputational capital more than the production of a large number of papers, each generating a small number of citations.

All of the productivity variables, which measure total publications as well as the quality of the research output, are highly significant and substantially improve the explanatory power of the model. As expected, introducing individual productivity measures in the model drives the coefficients on generalized experience and seniority toward zero. Given the model specification, the coefficients on LEVEL I ARTICLES and LEVEL II ARTICLES measure the differential effect of a top-tier or a second-tier publication over and above a publication in other outlets (the marginal effect of which is captured by the TOTAL PUBLICATIONS coefficient). It is clear that research output is highly rewarded in our sample of universities. (10) It is also clear that the market makes a distinction between the quantity and the quality of an individual's research output. Note that the marginal effect of a paper in a top-tier journal is significantly greater than the marginal effect of a paper in a second-tier journal, which in turn is gr eater than the marginal return to a publication in alternative outlets. (11)

The final measure of research quality, CAREER CITATIONS, provides additional evidence that an individual's relative earnings position depends on quality of his or her research output. While total career citations are valuable, the interaction term reveals an interesting trade-off between citations and the output of publications. For a given number of citations, the marginal reward for an additional citation is greater for an individual with fewer published articles than for an individual with more publications. Apparently, individuals who publish a smaller number of frequently cited papers add more to their reputational capital than individuals who publish a greater number of less frequently cited papers. (12) Nonetheless, the combined results on the reputational capital variables clearly indicate that there are significant monetary returns to high-quality research in the economics profession.

We also find a positive and significant return to experience as chairperson. Chairs in our sample receive almost a 2 percent premium for each year of service. Under this specification, the payoff to a chairperson with the sample average of 5.63 years of chair experience is about 9.6 percent; an estimate close to that reported from similar specifications in the literature. Finally, we find evidence that truly outstanding teaching pays off. For those individuals who have received at least one teaching award, the marginal return is about 7 percent.

How Durable Is Reputational Capital?

The profession is replete with examples of individuals who earn tenure and then quit publishing. At a less extreme level, it is not uncommon for some individuals to become relatively inactive during their careers. The question is, what--if any--penalties does an individual suffer when taking a research hiatus? Put differently, in the absence of additional investment, what is the decay rate on reputational capital associated with research inactivity?

To address this question, we construct the variable YEARS SINCE LAST LEVEL j = [D.sub.j] X [(93-[YR.sub.j]) + 1], where [YR.sub.j] denotes the year of the individual's last publication in journal level j and [D.sub.j] = 1 if the individual has at least one published paper in journal level j and zero otherwise. (13) For individuals whose most recent publication is in 1993 or for those individuals who have a forthcoming article in level j, YEARS SINCE LAST LEVEL j = 1. For individuals who have never published in level j, YEARS SINCE = 0. This construction ensures that since the total return from publishing level j articles for someone with no level j is zero, there can be no decay.

Table 3 reports the coefficients for variables measuring the years since an individual last published an article in a level I or level II journal. The conditional mean for years since publishing a level I article is about 8 years and about 5.5 years for a level II article. Both distributions, however, are skewed; over 50 percent of individuals in our sample published at least one level I article within the last 5 years, and a little over 66 percent published at least one level II article within the last 5 years. The longest publishing hiatus is 29 years for both journal levels. This is unusual, though. On net, individuals in our sample appear to be relatively active.

Table 3 reports the results from two alternative specifications. The first column contains the complete set of proxy variables for reputational capital. In this specification, the coefficients on both YEARS SINCE variables are insignificant, suggesting that reputational capital does not depreciate rapidly and that gaps in published research output do not lead to rapid reductions in an individual's relative earnings. The insignificant decay factor for these gaps may reflect the fact that an additional publication does not fully capture net increases in reputational capital. Apparently, even during a significant break in publishing, an individual's past publications continue to enhance reputational capital as citations to past research accumulate. In column (2), we examine this possibility by omitting citation variables from the model so that the publication variables pick up the combined salary effects of publications and their associated citations. While there continues to be no significant earnings penalty f or gaps in the production of level I papers, it appears that each year of research inactivity in level II journals leads to about a one-half percent reduction in relative earnings. The earnings decay associated with gaps in the production of level II may result from the substantial heterogeneity within this category. As a quality measure, level II publications have a larger variance than level I publications and thus, signal quality with more noise.

Relative Returns to Team Production

There has been a substantial increase in the incidence of co-authored papers over the last four decades. For example, there was a 440 percent increase in the proportion of co-authored papers appearing in seven leading U.S. journals between 1960 and 1990. (14) At the same time, the number of papers published per year in these journals increased by about 80 percent. In our sample, there was a total of 443 level I papers of which 65 percent were jointly produced. Of the 1009 level II articles 63 percent were jointly produced.

Does the market discount for joint work, and if so, how much? Our data permit us to directly address this question. Holding total publications constant, we separate the articles in the two high-quality levels according to whether they are single-authored or co-authored. (15) Table 4 reports regression coefficients on variables measuring the number of single- and co-authored papers an individual has written in each quality level. The estimates show that returns to papers published in the top-tier journals substantially exceed the returns to publishing in second-tier journals, whether single-authored or co-authored. While the relative magnitudes of the coefficients for level I and level II publications are as expected and bound the estimates in Table 2, it is interesting to note the slightly higher estimated returns to team production in both quality gradients. (16) Though we do not wish to make too much of these results, it may be worth noting the surprising conclusion in Diamond (1985): a citation to a multip le-authored article is worth more to its author than a citation to a single-authored article. Maybe, as Diamond speculates, team production represents a proxy for collegiality, which is rewarded by departments in the determination of salaries. Nevertheless, our results suggest that departments in our sample do not discount joint-authored papers. (17)

IV. SUMMARY AND CONCLUSIONS

The results from this study reinforce the notion that the quality of research plays a major role in determining an individual's reputation and hence, relative salary at research-oriented universities. It is clear that the market makes a distinction between the quantity and the quality of an individual's research output. That is, individuals who publish a smaller number of frequently cited papers add more to their reputational capital than individuals who publish a greater number of less frequently cited papers. Furthermore, reputational capital derived from the quantity and quality of published research tends to be durable and does not depreciate rapidly. This in turn implies that gaps in published research do not lead to rapid reductions in an individual's relative earnings. Even with a lengthy gap in an individual's publication stream, reputational capital can continue to grow as a result of the continuing accumulation of citations to past research output.

We also address a related question concerning the effect of team production on an individual's reputational capital. Our results suggest that the segment of the market examined here does not significantly discount joint-authored papers. These results represent an interesting counter-point to those of Sauer (1988). His estimates suggest that the weight for co-authored papers is not much different from 1/n; the market fully discounts the returns to multiple-authored papers by the number of authors involved. The difference between our estimates and Sauer's may merely reflect the different populations from which the samples were drawn. Sauer's sample frame consisted of economists from seven of the top 40 departments, while our sample consists primarily of departments below the top 40. While these two studies may represent samples drawn from two non-competing markets, this begs a question. Why would one market fully discount research derived from team production, while the other does not? One very tentative explan ation may be that joint research at top programs reflects the combination of substitute-skilled individuals. On the other hand, research output generated at other programs involves the combination of complementary-skilled individuals. Resolution of this question, however, remains outside the scope of this article.

Moore: Gulf Coast Coca-Cola Professor, Dept. of Economics, Louisiana State University, Baton Rouge, LA 70803. Phone 1-225-578-3792, Fax 1-225-578-3807, E-mail [email protected]

Newman: South Central Bell Professor, Dept. of Economics, Louisiana State University, Baton Rouge, LA 70803. Phone 1-225-578-3794, Fax 1-225-5783807, E-mail [email protected]

Turnbull: Professor, Dept. of Economics, Georgia State University, Atlanta, GA 30303. Phone 1-404-651-2626, Fax 1-404-651-4985, E-mail [email protected]

(*.) We wish to thank two anonymous referees for helpful comments and suggestions.

(1.) See McDowell and Melvin (1983), Barnett et al. (1988), Laband and Piette (1995), and Mixon (1997).

(2.) We define the top programs as those programs rated highly in studies by Moore (1972), Siegfried (1972), Bell and Seater (1978), Graves et al. (1982), Hogan (1986), and Scott and Mitas (1996). This screening strategy resulted in a total of 13 pedigree programs: University of Chicago, Harvard University, Massachsetts Institute of Technology, University of Michigan, University of Minnesota, Northwestern University, University of Pennsylvania, Princeton University, Stanford University, University of California, Berkeley, University of California, Los Angeles, University of Wisconsin, and Yale University. A more restrictive list including only Chicago, Harvard, MIT, Yale, Pennsylvania, and Stanford does not affect the results reported below.

(3.) Our choice of top 10 journals includes the top 8 used in the recent department ranking study by Conroy et al. (1995), plus the Economic Journal and Economica.

(4.) Our list of the top 55 journals also overlaps significantly with the list of 36 journals used by Scott and Mitias (1996) in their ranking of economics departments.

(5.) See Hamermesh et al. (1984), Diamond (1986), and Sauer (1988).

(6.) Alternative empirical models using a dummy variable to capture chair experience yield the same qualitative results as those reported for the chair variable defined here.

(7.) The existing evidence is mixed, however. See Siegfried and white (1973), Tuckman and Leahey (1975), Katz (1978), and Ferber (1974).

(8.) We include only associate and full professors in this study. Many of the issues addressed below apply almost exclusively to individuals who have had sufficient time in the profession to establish their reputational capital. Also, we find that the reward structure for assistant professors is significantly different from that of tenured faculty. Experience and seniority, for instance, are almost perfectly collinear for the assistant professors in our sample.

(9.) See Gordon et al. (1974), Hoffman (1976), Ransom (1993), and Moore et al. (1998) for detailed analysis of this particular result.

(10.) Introducing quadratic publication variables into the model yields estimates consistent with diminishing returns, as reported by Tuckman and Leahey (1975) and Hansen et al. (1978). These additional variables do not alter our conclusions regarding research hiatus and co-authoring effects, so we report the simple linear formulations below.

(11.) The basic model was also estimated with dummy variables identifying each individual's primary area of specialization: econometrics, money, general economics, development, international, industrial organization, agricultural economics, labor, and urban/regional. None of the included dummy variables was significant at conventional levels of significance.

(12.) Alternative specifications included additional measures of citation concentration. We introduced separately variables measuring the total number of citations to each individual's top 1, 2, 3, and 4 articles. Additionally, we examined the effect of the proportion of an individual's research output that received at least one citation. The addition of these "concentration" measures did not change the qualitative results, but they did somewhat complicate the interpretation of the partial derivatives. Therefore, we opted for the more straight forward interactive term, which permits us to address the concentration hypothesis with a sample of researchers exhibiting substantial variation in the number of published papers. The interaction term has intuitive appeal because the estimated marginal effect of a citation is allowed to vary with the individual's stock of publications; that is, as a measure of quality, it recognizes the multiplicative relationship between the number of publications and career citations.

(13.) Recall that the terminal year in our sample is 1993.

(14.) These are: American Economic Review, Journal of Political Economy, Quarterly Journal of Economics, Review of Economics and Statistics, Economic Inquiry, Econometrica, and Southern Economic Journal.

(15.) Unlike Sauer (1988), we do not distinguish between papers with 2 or more authors. Our objective is merely to determine whether single-authored papers generate larger marginal returns to individuals than co-authored papers, irrespective of the exact number of authors.

(16.) The parameter values within each publication category are not statistically different from one another: level I ([F.sub.2,117] = 0.188) and level II ([F.sub.2,117] = 0.942).

(17.) One referee suggested that since single-authored papers may be relatively abundant in earlier periods and relatively scarce in later periods, our specification confounds the effect of time since publication with the effect of joint authorship. A formal examination of this issue, however, would require more data than we have available (including panel data on salaries).

REFERENCES

Barnett, A., R. Ault, and D. Kaserman. "The Rising Incidence of Co-Authorship in Economics: Further Evidence." Review of Economics and Statistics, LXX(3) 1988, 539-43.

Bell, J. G., and J. J. Seater. "Publishing Performance: Departmental and Individual." Economic Inquiry, 16, 1978, 599-615.

Conroy, M., R. Dusansky, D. Crukker, and A. Kiledgaard. "The Productivity of Economics Departments in the U.S.: Publications in Core Journals." Journal of Economic Literature, 33, 1995, 1966-71.

Diamond, A. M., Jr. "The Money Value of Citations to Single-Authored and Multiple-Authored Articles." Scientometrics, 8(5-6), 1985, 315-20.

-----. "An Economic Model of the Life-Cycle Research Productivity of Scientists." Scientometrics, 6(3), 1984, 189-96.

-----. "What is a Citation Worth?" Journal of Human Resources, 21, 1986, 200-15.

Ferber, M. "Professors, Performance, and Rewards." Industrial Relations, 13(1), 1974, 69-77.

Gordon, N. M., T. E. Morton, and I. Braden. "Faculty Salaries: Is There Discrimination by Sex, Race, and Discipline?" American Economic Review, 64(3), 1974, 419-27.

Graves, P. E., J. R. Marchand, and R. Thompson. "Economics Departmental Rankings: Research Incentives, Constraints, and Efficiency." American Economic Review, 71(5), 1982, 1131-41.

Hamermesh, D., G. E. Johnson, and B. A. Weisbrod. "Scholarship, Citation, and Salaries: Economic Rewards in Economics." Southern Economic Journal, 49(2), 1984, 472-81.

Hogan, T. D. "The Publishing Performance of U.S. Ph.D. Programs in Economics During the 1970s." Journal of Human Resources, 21(2), 1986, 216-29.

Katz, D. A. "Faculty Salaries, Promotions and Productivity at a Large University." American Economic Review, 63, 1978, 469-77.

Laband, D. N., and M. J. Piette. "Team Production in Economics: Division of Labor or Mentoring?" Labour Economics, 2, 1995, 33-40.

Liebowitz, S. J., and J. P. Palmer. "Assessing the Relative Impacts of Economic Journals." Journal of Economic Literature, 22, 1984, 77-88.

McDowell, J., and M. Melvin. "The Determinants of Co-Authorship: An Analysis of the Economics Literature." Review of Economics and Statistics, LXV(1), 1983, 155-60.

Mixon, F., Jr. "Team Production in Economics: A Comment and Extension." Labour Economics, 4, 1997, 185-91.

Moore, W. J. "The Relative Quality of Economic Journals: A Suggested Rating System." Western Economic Journal, 10(2), 1972, 156-69.

Moore, W. J., R. J. Newman, and G. K. Turnbull. "Do Academic Salaries Decline with Seniority?" Journal of Labor Economics, 16(2), 1998, 352-66.

Ransom, M. R. "Seniority and Monopsony in the Academic Labor Market." American Economic Review, 83, 1993, 221-33.

Saks, D. H. "How Much Does a Department Chairperson Cost?" Journal of Human Resources, 12(4), 1977, 535-40.

Sauer, R. D. "Estimates of the Returns to Quality and Coauthorship in Economic Academia." Journal of Political Economy, 96(4), 1988, 855-66.

Scott, L. C., and P. M. Mitias. "Trends in Rankings of Economics Departments in the U.S.: An Update." Economic Inquiry, 34, 1996, 378-400.

Siegfried, J. J. "The Publishing of Economic Papers and Its Impact on Graduate Faculty Ratings." Journal of Economic Literature, 10(1), 1972, 31-73.

Siegfried, J. J., and K. J. White. "Financial Rewards to Research and Teaching: A Case Study of Academic Economists." American Economic Review, 63(2), 1973, 309-15.

Tuckman, H. P., and J. Leahey. "What Is an Article Worth?" Journal of Political Economy, 83(5), 1975, 951-67.

TABLE 1

Characteristics of Faculty at Nine Ph.D. Programs in Economics,
Means and Standard Deviations

 Full Associate
Variable Professor Professors

SALARY ($) 72,897 51,986
 (18,962) (7,633)
EXPERIENCE 21.86 13.71
 (6.10) (5.58)
SENIORITY 16.49 11.91
 (7.77) (6.27)
TOTAL PUBLICATIONS 45.41 16.47
 (34.91) (8.53)
LEVEL I ARTICLES 3.74 2.09
 (4.24) (1.96)
LEVEL II ARTICLES 10.02 4.16
 (8.36) (3.04)
CAREER CITATIONS 147.00 28.91
 (190.00) (33.26)
TEACHING AWARD .21 .11
MALES .98 .93
FOREIGN .05 .11
PH.D. PEDIGREE .40 .42
TABLE 2

Determinants of Earnings for Tenured Faculty

Variable Model (1) Model (2)

Demographic Characteristics:

 EXPERIENCE .059 (*) .011
 (4.31) (1.14)
 EXPERIENCE (2) -.001 (*) -.0001
 (2.54) (0.91)
 SENIORITY -.024 (*) -.001
 (7.28) (0.44)
 SEX (Male = 1) .118 .026
 (1.17) (0.40)
 PH.D. PEDIGREE .020 .032
 (0.26) (1.14)
 FOREIGN .021 .014
 (0.51) (0.26)

Reputational Capital:

 TOTAL PUBLICATIONS .001 (*)
 (2.24)
 LEVEL 1 ARTICLES .028 (*)
 (5.78)
 LEVEL II ARTICLES .016 (*)
 (6.27)
 CAREER CITATIONS .001 (*)
 (3.65)
 CITATIONS (*) TOTAL ARTICLES -.00002 (*)
 (5.36)

Administrative Returns:

 YEARS AS CHAIR .017 (*)
 (3.27)

Return to Teaching:

 TEACHING AWARD .069 (**)
 (1.95)
Intercept (a) 10.417 (*) 10.541 (*)
 (61.27) (93.71)
[R.sup.2] .35 .73
Sample size 140 140

Note: Dependent variable is log of 9-month salary; [absolute value t]
statistics are in parentheses.

(a)Deparptment dummy variables are not reported.

(*)Significant at the .05 level.

(**)Significant at the .10 level.
TABLE 3

Durability of Reputational Capital

Variable Mean (1) (2)

TOTAL PUBLICATIONS .001 (*) .003
 (2.10) (0.47)
LEVEL I ARTICLES .028 (*) .026 (*)
 (5.72) (5.54)
LEVEL II ARTICLES .016 (*) .013 (*)
 (6.04) (4.78)
CAREER CITATIONS .001 (*)
 (3.39)
CITATIONS (*) TOTAL ARTICLES -.00002 (*)
 (4.78)
YEARS SINCE LAST LEVEL I 8.06 (a) .0004 .002
 (7.02) (0.19) (0.96)
YEARS SINCE LAST LEVEL II 5.46 (a) -.002 -.006 (*)
 (5.31) (0.84) (2.22)
[R.sup.2] .72 .67

Note: Also included in the regression are the other variables listed in
model (2), Table 2.

(a)Conditional on at least one publication at this quality level.

(*)Significant at .05 level.
TABLE 4

The Effects on Earnings of Single Authored and Co-authored
Articles

Variable Coefficient t-value

LEVEL I ARTICLES--SINGLE .025 (*) 2.77
LEVEL I ARTICLES--JOINT .029 (*) 4.54
LEVEL II ARTICLES--SINGLE .012 (*) 2.46
LEVEL II ARTICLES--JOINT .017 (*) 5.83

Note: Also included in the regression are the other variables listed in
Model (2), table 2.

(*)Significant at the .05 level.

APPENDIX: Level II Journals

1. American Economic Association Papers and Proceedings

2. American Journal of Agricultural Economics

3. Brookings Papers on Economic Activity

4. Canadian Journal of Economics

5. Econometric Theory

6. Economic History Review

7. Economic Development and Cultural Change

8. Economic Inquiry

9. History of Political Economy

10. Industrial and Labor Relations Review

11. Journal of Business

12. Journal of Business and Economic Statistics

13. Journal of Comparative Economics

14. Journal of Econometrics

15. Journal of Economic Dynamics and Control

16. Journal of Economic History

17. Journal of Economic Literature

18. Journal of Economic Behavior and Organization

19. Journal of Environmental Economics and Management

20. Journal of Finance

21. Journal of Financial Economics

22. Journal of Financial and Quantitative Analysis

23. Journal of Health Economics

24. Journal of Human Resources

25. Journal of Industrial Economics

26. Journal of Institutional and Theoretical Economics

27. Journal of International Money and Finance

28. Journal of International Economics

29. Journal of Labor Economics

30. Journal of Law and Economics

31. Journal of Legal Studies

32. Journal of Money Credit and Banking

33. Journal of Macroeconomics

34. Journal of Mathematical Economics

35. Journal of Monetary Economics

36. Journal of Public Economics

37. Journal of Regional Science

38. Journal of the American Statistical Association

39. Journal of Urban Economics

40. Kyklos

41. National Tax Journal

42. Oxford Economic Papers

43. Public Choice

44. Rand Journal of Economics

45. Southern Economic Journal