Academic achievement by graduates from for-profit and nonprofit institutions: Evidence from CPA exam performance

ABSTRACT

This study shows that graduates from nonprofit educational institutions outperform graduates from for-profit institutions on the four sections of the certified public accountant (CPA) exam. Specifically, it (1) documents univariate differences in CPA exam scores, score distributions, pass rates, and time to complete the CPA exam; (2) investigates the differences in a multivariate context; and (3) examines the robustness of the results to various levels of control for student differences.

KEYWORDS:

• Accounting academic achievement
• CPA exam performance
• educational institution structure
• for-profit versus nonprofit

Acknowledgments

We wish to thank Alfonzo Alexander, Olaf Wasternack, and James Suh from NASBA for their assistance with obtaining and analyzing the CPA exam data. We also acknowledge helpful comments from presentation participants at two 2014 NASBA regional meetings and the 2014 American Accounting Association annual meeting, along with valuable suggestions from anonymous reviewers.

Funding

Professor Mittelstaedt gratefully acknowledges the financial support of the Deloitte Foundation.

Notes

1 For more information on the CPA exam, see http://www.aicpa.org/BECOMEACPA/CPAEXAM/Pages/CPAExam.aspx.

2 Nonprofit institution codes are 501(c)3, Exempt, or 170(c)1. For-profit institutions have a code of Proprietary.

3 The for-profit/nonprofit designations are clear, but recent NASBA presentations have revealed that their database may misclassify observations regarding undergraduate and graduate designations. As a result, the graduate/undergraduate analysis was not conducted in this study.

4 The sections have different content, and we are unaware of any causal links between passing one section (e.g., AUD) and passing another section (e.g., REG).

5 “Best Colleges Rankings” can be viewed at the following website: http://premium.usnews.com/best-colleges. The ranking formula uses up to 16 quantitative measures that education experts have proposed as reliable indicators of academic quality.

6 As mentioned earlier in the paper, each record in the database has an IRS code to designate the tax status (for-profit/nonprofit) of the institution from which they were graduating. If a candidate obtained more than one degree and the tax status did not match, the observation was eliminated from the study.

7 Because each multiple choice question included four possible answers and 60% was the lowest percentage of multiple choice questions on any section of the CPA exam, individuals would be expected to achieve a minimum SectionScore of 15 by guessing (without answering any simulation questions). As a result, SectionScores below 15 were judged to be frivolous attempts by the candidate and were eliminated from the analysis.

8 These control variable results are consistent with a working paper (Trinkle et al., 2015) examining the effect of age and gender on CPA examination performance.

9 Each observation in the database indicated their degree type (BA, MBA, etc.). The observations selected for this analysis had degree type equal to “NONE.”

10 Since there was no degree date available for these observations, the condition to be within −1/2 year and 2 years of graduation, which was required in all other samples used in this study, could not be implemented.