Skip to main content

Table 9 Replication of Tables 1 and 3 with author fixed effects

From: Journal rankings and publication strategy

  ABS ranking change SJR indicator change quartile
  Constant Downgraded Upgraded Bottom 25% Middle 50% Top 25%
  (1) (2) (3) (4) (5) (6)
ABS15 0.0963 -0.2497* -0.1580 0.2381 0.0364 -0.4433*
  (0.2280) (0.0782) (0.1955) (0.1370) (0.1826) (0.1571)
2nd Version 7.3646 0.5562* 1.7734 3.3925* 4.1500 2.2769
  (3.3749) (0.2264) (0.7703) (1.4117) (2.0242) (1.1343)
3rd Version 12.2334* 0.7173*** 3.2370** 4.8999** 8.4121** 2.7454*
  (3.9877) (0.0660) (0.9135) (1.3860) (2.1394) (0.9623)
4th and subsequent 15.4312** 0.8634** 6.8384*** 4.2382** 14.3618*** 4.1853***
Version (2.9186) (0.2220) (0.9771) (1.0311) (2.1791) (0.4099)
Observations 51,021 51,021 51,021 51,021 51,021 51,021
R2 0.1387 0.0616 0.0893 0.1034 0.1045 0.0749
  1. Notes: Dependent variables are publications in journals in a category described in the column title. Constant refers to journals that retained their ABS journal ranking between 2010 and 2015, While Downgraded and Upgraded refers to journals that had their rankings decreased and increased, respectively. The variable takes value 100 if a paper is published in a journal of that category and zero otherwise. ABS15 is a dummy variable that takes value one in years in which the ABS15 ranking is valid for the UK-based scholars, i.e. years 2015, 2016 and 2017. Variables 2nd—4th and subsequent version are dummy variables that represent the number of the working paper versions uploaded to the repository. We cap the number of versions at 4.
  2. All regression specifications include paper age and author effects and the number of versions covariates. For papers with multiple authors, each observation is replicated and weighted accordingly. All standard errors are clustered at the age level. *** p < 0.001, ** p < 0.01, * p < 0.05.