A burden of knowledge creation in academic research: evidence from publication data

ABSTRACT

Academic research is vital for innovation and industrial growth. However, a potential burden of processing ever more knowledge could be affecting research output and researchers’ careers. We look at a dataset of researchers who have published in journals in the field of economics during a period of 45 years. For a subset of these researchers, we amass data from journals listed in the EconLit database, supplemented with years of birth from public sources. Our results show an increase in the age of researchers at their first publication, in the number of articles referenced in debut articles, and in the number of co-authors. Simultaneously, we observe a decline in the probability of researchers changing research fields. Our findings extend earlier findings on patents and hint at a burden of knowledge pervading different areas of human progress. Moreover, our results indicate that researchers develop strategies of specialisation to deal with this challenge.

KEYWORDS:

• Knowledge creation
• academic research
• researcher age
• team size
• references
• specialisation

Disclosure statement

No potential conflict of interest was reported by the authors.

Supplementary Material

Supplemental data for this article can be accessed here.

Notes

¹ This is also argued by Segerstrom (1998). An assumption in this context is that problems are independent. The effect in the case of problems with sequential dependencies is unclear, since there is no obvious reason to expect later problems to be easier (or harder) to solve than earlier problems.

² It should be noted that Conley et al. (2013) correct for the number of authors and quality measured in AER equivalents. Absent those corrections, differences are less pronounced.

³ The oldest professional journal of economics is the Quarterly Journal of Economics which appeared first in 1886.

⁴ The dynamics of the journal landscape pose a difference to the field of patents where patent offices and categories stay more persistent over time.

⁵ We chose this list of top-ranking peer-reviewed journals after considering several different rankings, including Ritzberger (2008), Engemann and Wall (2009), Kodrzycki and Yu (2006), and www.journal-ranking.com. An overview of the journals and their numbers of articles, authors and time periods covered is provided in Table A1 in Appendix A (The Electronic Appendix is available at http://www.electronic-appendix.info/bok). Other studies have focused on similar samples of top-ranking journals, such as Kim, Morse, and Zingales (2006), Hamermesh (2013), or Card and DellaVigna (2013).

⁶ The Electronic Appendix is available at http://www.electronic-appendix.info/bok.

⁷ We elaborate on the data sources used in this study in Appendix B (The Electronic Appendix is available at http://www.electronic-appendix.info/bok).

⁸ The Electronic Appendix is available at http://www.electronic-appendix.info/bok.

⁹ There are combinations of journals and years with no reference data available. To check for a potential bias of journals with higher numbers of references substituting those with lower numbers, we conducted separate regression analyses for each journal where data is available for at least 100 articles over at least 10 years. In 98.4% of the 124 journals that we detect a significant time trend for, the trend is positive.

¹⁰ The Electronic Appendix is available at http://www.electronic-appendix.info/bok.

¹¹ We thank an anonymous referee for pointing this out.

¹² To counter the inflation of quantity of research papers, potentially at the cost of quality, it has been proposed to focus on the citations per article, instead of focusing on the size of a researcher’s list of publications when tenuring (e.g. Hirsch 2005; Egghe 2006; Marx 2011; Ellison 2013).

¹³ Despite an increase in life expectancy, Hazan (2009) finds that expected hours of lifetime labour supply have decreased. Lee (2001, 648) estimates the length of male retirement in the U.S. to have increased by more than factor six since 1850, representing ‘up to 30% of men’s length of life after entry into the labor force’.