39
2019
TRENDS REPORT
SECTION 07
ENDNOTES
a. The statistical vendor changed from CXACT to ARC Research between the 2015 and 2017 reports. ARC did their due diligence in researching
methods from previous years and creating a model that is as comparable as possible to previous years’ data. The 2016 and 2018 models are based
on a linear regression on weighted data, using faculty size as a predictor. For 2016 and 2018, weights were applied across the whole survey sample.
For 2014, the previous vendor noted that separate weights were calculated for Bachelor’s, Master’s, and Ph.D. respondents.
b. For 2016 and 2018, the weighted proportions of MA program types were applied to the projected MA total to derive breakdowns by program type.
For 2014, it appears that the unweighted proportions of MA program types were used. For 2012 and years prior, it is unclear how proportions of MA
program types were derived.
c. Enrollment data collected for the 2015 Tr ends report represents the 2014-15 academic year instead of the 2013-14 academic year due to the wording
of the supply survey questionnaire for this period.
d. The statistical vendor changed from CXACT to ARC Research between the 2015 and 2017 reports. ARC did their due diligence in researching
methods from previous years and creating a model that is as comparable as possible to previous years’ data. The 2016 and 2018 models are based
on a linear regression on weighted data, using number of CPAs to predict total new graduate hires with accounting degrees. For 2016 and 2018, the
weighted proportions of BA vs. MA were applied to the projected total to derive breakdowns by degree level. For 2014, it appears that the unweighted
proportions of BA/MA were used. For 2012 and years prior, it is unclear how proportions of BA/MA hires were derived.
e. To avoid double counting of dual degree students, respondents were instructed to classify students as follows:
01. Enrollments - Respondents were instructed to classify students in years 1 through 4 as Bachelor’s and students in year 5 as Master’s.
02. Graduates - Respondents were instructed to exclude dual degree graduates from the Bachelor’s totals, and include them in the Master’s totals.
f. Data collection for the multiethnic race/ethnicity category began in 2010.
g. In 2014 and previous years, areas of assignment were limited to Accounting/Auditing, Taxation, MIS/Computers, and Other. In 2016, we broadened
the areas for the snapshot data (3.2), but kept the condensed categories for the trends data (3.7) as follows:
Additionally, this year we added an “Internal Audit” category for further segmentation.
AREAS OF ASSIGNMENT CONDENSED CATEGORY
Financial Accounting Acc./Aud.
Auditing, not Including Internal Audit Acc./Aud.
Internal Audit Acc./Aud.
Assurance Services Other
Taxation Taxation
Information Management and Technology Assurance Other
Consulting Other
Financial Forensics Other
Business Valuation Other
Personal Financial Planning Other
Management Accounting Other
Transaction Services Other
Other Areas Other
ENDNOTES
h. For this chart, 2014 race/ethnicity fi gures have been recalculated to exclude “Unknown” for comparability with other years.
i. In previous years, data on the hiring of non-accounting degree holders only included “non-accounting business majors” and “non-business
majors.” In 2018, we changed the categories to “non-accounting business majors – e.g., fi nance, marketing, etc.,” “STEM majors, not
including accounting or business,” and “all other majors.” Because of the addition of more specifi c categories, the total of non-accounting
new hires is likely artifi cially infl ated in 2018 and isn’t directly comparable to previous years. These new categories are indicative of trends
in the public accounting industry and are crucial in tracking industry trends accurately in the future.
j. In iterations of the report prior to 2016, the data in this chart included all fi rms that responded to the survey, rather than only fi rms that hired
in the survey year. Thus, some percentage of those in previous years that noted that they would hire “the same” next year as in the survey
year did not hire in that year and did not plan to hire in the future year (these tend to be smaller fi rms). This representation of the data was
improved upon and the charts in 2017 and 2019 only include prospective hiring of fi rms that hire.
k. Due to small sample size, this data cannot be generalized to the entire fi rm universe. This data, instead, is simply a report of the fi rms that
actually responded.
l. Firm demographics are not broken down by fi rm size due to the insuffi cient samples in the 50-200 and 200+ CPA categories. The number
of respondents in those categories is too small (i.e., there are fewer than 40 total respondents in each of those categories) to draw valid
conclusions from despite relatively higher response rates in those categories.
m. CPA Exam data often show increases in candidates prior to changes to the Exam followed by signifi cant decreases in the fi rst few years
after the CPA Examination changes. With the changes in 2017, decreases were expected in 2018.