In this section, we consider four main...

In this section, we consider four main measures of impact: Google Scholar citations, SSRN downloads, citations in SEC rules, and media coverage. Citations are the most traditional measure of academic impact, whereas SSRN downloads can be high for papers that are read more broadly than people just in the field. SEC citations are collected from proposed and final rules following and extending the sample of Geoffroy and Lee (2021).^² Panel A of Figure 2.4 plots five-year rolling average citations per paper for forensic and non-forensic papers by publication year from 2004 to 2022. Forensic papers have citation rates that are almost identical to other papers. To assess statistical significance, Table 2.1 reports results for a regression of citations on an indicator for forensic papers controlling

We use data provided by Geoffroy and Lee (2021) prior to 2017 and extend their sample from 2018 to present. They show that there were few citations before 2011 when the court’s 2011 ruling against the SEC in the Business Roundtable v. SEC (647 F. 3d 1144, Court of Appeals, D.C. Cir. 2011) drastically increased the citations of academic papers in proposed rulemaking and resulted in fewer negative comment letters. SSRN downloads may be undercounted to the extent that working paper versions carry different titles and are not linked in SSRN to the published version. Additionally, some authors post versions on their websites or other places and not SSRN.

Figure 2.4

A line graph showing Average Google Scholar Citations by Paper Publication Year from 2004 to 2022 for Forensic and Others papers. A line graph showing average SSRN downloads by paper publication year from 2004 to 2022 for Forensic and Others papers.

View large Download slide

Google Scholar citations and SSRN downloads.

Notes: This figure shows the comparison of Google Scholar citations and SSRN downloads between forensic finance and other papers. A paper is categorized as being forensic finance if it uses forensic words at least 20 times or at least 10 times with at least one usage in the title or abstract. In Panel A, it shows the 5-year rolling average Google Scholar citations of forensic finance (other) papers. The full sample is used, which includes 6334 published and forthcoming papers in the top three finance journals between 2000 and April 2023. In Panel B, it shows the 5-year rolling average SSRN downloads of forensic finance (other) papers. The sample includes 5142 papers (81% of full sample) which have posted a working paper version on SSRN. In both panels, red represents forensic finance papers and grey represents all other papers in the sample. Both Google Scholar citations and SSRN downloads are collected in 2023.

Table 2.1

Citation regressions

	Google Scholar		SSRN		Press		SEC
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
Forensic	38.926	37.454	125.916**	163.412**	0.303**	0.315***	0.069***	0.069***
Forensic	(24.562)	(23.606)	(59.475)	(60.757)	(0.094)	(0.094)	(0.019)	(0.021)
JEL FE		✓		✓		✓		✓
Journal FE	✓	✓	✓	✓	✓	✓	✓	✓
Year FE	✓	✓	✓	✓	✓	✓	✓	✓
Observations	6,334	5,423	5,142	4,768	2,816	2,642	6,334	5,423
R²	0.310	0.340	0.035	0.066	0.017	0.019	0.021	0.029
Dep. Var. Mean	299.134	286.841	1001.224	1014.602	0.454	0.455	0.079	0.083

	Google Scholar		SSRN		Press		SEC
	(1)	(2)	(3)	(4)	(5)	(6)	(7)	(8)
Forensic	38.926	37.454	125.916**	163.412**	0.303**	0.315***	0.069***	0.069***
Forensic	(24.562)	(23.606)	(59.475)	(60.757)	(0.094)	(0.094)	(0.019)	(0.021)
JEL FE		✓		✓		✓		✓
Journal FE	✓	✓	✓	✓	✓	✓	✓	✓
Year FE	✓	✓	✓	✓	✓	✓	✓	✓
Observations	6,334	5,423	5,142	4,768	2,816	2,642	6,334	5,423
R²	0.310	0.340	0.035	0.066	0.017	0.019	0.021	0.029
Dep. Var. Mean	299.134	286.841	1001.224	1014.602	0.454	0.455	0.079	0.083

Notes: This table examines the difference in four main measures of impact between forensic finance and other papers. We estimate the OLS regression of the form:

Citation_i = α + βForensic_i + FEs + ϵ_i,

where Forensici is a dummy variable equal to 1 if paper i is flagged as a forensic finance paper. A paper is categorized as being forensic finance if it uses forensic words at least 20 times or at least 10 times with at least one usage in the title or abstract. Fixed effects are indicated at bottom of each column. As for JEL fixed effects, we assign each paper a unique JEL indicator (non-G, G1, G2, G3, or other-G) based on the paper’s most common 1-digit and 2-digit JEL codes. The full sample includes 6334 published and forthcoming papers in the top three finance journals between 2000 and April 2023. There are 5142 (81%) papers for which we found a working paper version on SSRN. For the Journal of Financial Economics and recent papers in the Review of Financial Studies, we collect JEL codes from papers’ published versions. For the Journal of Finance and earlier papers in the Review of Financial Studies, we collect JEL codes from working paper versions of published papers, which are available for 69% of papers. All four measures of impact are collected in 2023. Press citations are obtained from Altmetric and include posts from various news outlets. SEC citations are collected from SEC proposed rules and SEC final rules released between 2007 and 2022. All dependent variables are winsorized at 95% percentile. Standard errors are clustered by year and reported in parentheses. ***p < 0.01, **p < 0.05, *p < 0.1.

View large

for publication year and journal fixed effects, as well as the paper’s most common one or two digit JEL code. On average, forensic papers have an extra 39 citations per paper, but this is statistically insignificant and economically modest relative to the mean of 299 citations per paper. Panel B of Figure 2.4 repeats the same exercise for SSRN downloads, which is another potential measure of impact. Forensic downloads vary somewhat year-to-year but are higher than non-forensic papers in most years. Regressions in Table 2.1 show forensic papers to have an average of 126 more downloads, which is 13% of the mean and significant at the 5% level.

Sharing Unavailable