Economists' perceptions of their own research: a survey of the profession

American Journal of Economics and Sociology, The,  April, 1997  by William L. Davis

• E-mail
• Print
• Link

I

Introduction

Professional economists from various backgrounds are often asked to participate in opinion surveys about the economy, economic and social issues, and politics. Studies by Alston, et al. (1992). and Rickets, et al. (1992). which attempted to measure the degree of consensus among American economists and British economists respectively on various economic issues are among the most recent surveys. In these surveys, economists were asked to comment on the efficiency of resource allocation or the optimality of government policy. Rarely, however, have economists been polled on their perceptions of the social value of their own scientific research.

During the last several years, the economics profession has been criticized for its emphasis on abstract theoretical exposition and its seeming lack of ability and/or willingness to address real-world economic issues. The profession has been criticized for the methods used in the training of professional economists in the United States as well. The professional training of economists at many leading American universities may entail numerous hours of esoteric reading and require substantial mathematical competence. The following quote from the Wall Street Journal illustrates this criticism:

Graduate programs may be turning out a generation with too many idiot savants, skilled in technique but innocent of real economic issues. At one unnamed "leading" university, graduate students couldn't figure out why barbers' wages have risen over time, "but they could easily" solve a two-sector general equilibrium model with disembodied technical progress in one sector (Vogel, 1995).

Recently, a new field within the discipline, feminist economics, is being developed by individuals within the profession who believe that economic theory and evidence is driven by its large male population. The Committee on the Status of Women in the Economics Profession reports that only nine percent of academic economists in 1989 were women. The feminist approach in economics asserts that the profession has largely ignored the role, behavior, and progress of women and many noncaucasion groups in the economy. Those in Feminist Economics are attempting to expand the theoretical framework used by conventional economists to be more reflective of the state of the world (Ferber, 1995; Blank, 1993).

As a result of the various criticisms of the profession, it seemed appropriate to survey the profession as part of an effort to ascertain economists' views of their work. This paper presents the results of such a survey conducted by mail among 1500 economists.

II

Design of the Survey Instrument

The survey questionnaire consisted of 28 general propositions which reflect anecdotal and/or empirical criticisms about various aspects of the economics profession. Before its delivery, the questionnaire was pre-tested among a limited number of economists. Many suggestions from the pre-test, including the deletion of some and addition of other propositions, were incorporated into the survey in an effort to avoid biased and/or unclear interpretations by those responding.

The questionnaire was mailed to economists living in the U.S. and Canada randomly selected from the American Economic Association's 1993 membership directory, a comprehensive directory of professional economists. Following Alston, et al. (1992) and Kearl, et al. (1979). respondents were asked to agree, agree with caveats, or disagree with each proposition as a factual statement. All propositions were phrased as positive (as opposed to normative) statements. During a five-week period, 734 responses were received - a response rate of 49 percent. The response rate compares favorably with those of the Alston, et al., Ricketts, et al., and Kearl, et al. studies, each of which generated approximately 35 percent.

A factor analysis with the factor loadings criteria set to four was performed on the initial survey data, confirming four predetermined additudinal dimensions. Factor loadings ranged from .36 to .83 and eigenvalues ranged from 1.4 to 5.2 that explained 39 percent of the variance in the responses.

Based on the factor analysis, the propositions were placed in one of four different categories reflecting the respondents': (1) perceptions about research methodology, (2) perceptions about the usefulness and social value of economic research, (3) perceptions about the publication of that research, and (4) perceptions about the influence of race and gender on economic research. Also following the work of Alston, et al. the degree of consensus for each proposition was estimated using a relative entropy index. The entropy index ranges between zero and one; high entropy values imply less consensus and vice versa. The relative entropy index equals one if the respondents are distributed equally across all three categories. The index would equal zero if all respondents chose the same category, implying a perfect consensus.

Table 1

Background Characteristics of Total Respondents

Percentage of Respondents That:

were
Male                                           85
Female                                         15

worked in
Academia                                       84.7
Business and Industry                           3.7
Government                                      6.1

completed his/her economics degree during
1941-70                                        26.7
1971-80                                        32.5
1981-90                                        32.2
1991-93                                         8.7