Table A2
ANOVA table, comparing mean audit fees across different years (2015–2020)
| N companies = 269 | F-value | p-value |
|---|---|---|
| (Intercept) | F(1,1242) = 40128.00 | <0.01 *** |
| TENURE | F(1,1242) = 6.84 | 0.01 ** |
| ASSETS | F(1,1242) = 1421.32 | <0.01 *** |
| INVREC | F(1,1242) = 6.52 | 0.01 ** |
| LAG | F(1,1242) = 26.03 | 0.00 *** |
| BUSY | F(1,1242) = 3.83 | 0.05 ** |
| ROA | F(1,1242) = 0.40 | 0.53 |
| BIG4 | F(1,1242) = 3.78 | 0.05 ** |
| Year-end (related to COVID 19 extension) | F(1,1242) = 2.91 | 0.09 |
| Year | F(5,1242) = 21.58 | <0.01 *** |
| Year*BUSY | F(5,1242) = 1.65 | 0.14 |
| Year*BIG4 | F(5,1242) = 1.16 | 0.33 |
| Year*Year-end (related to COVID 19 extension) | F(5,1242) = 2.64 | 0.02 ** |
| Year*TENURE | F(5,1242) = 3.21 | 0.01 ** |
| BUSY*BIG4 | F(1,1242) = 11.31 | 0.00 *** |
| BUSY*Year-end (related to COVID 19 extension) | F(1,1242) = 2.75 | 0.10 |
| BUSY*TENURE | F(1,1242) = 1.93 | 0.17 |
| BIG4*Year-end (related to COVID 19 extension) | F(1,1242) = 0.28 | 0.60 |
| BIG4*TENURE cat | F(1,1242) = 1.12 | 0.29 |
| Year-end (related to COVID 19 extension)*TENURE | F(1,1242) = 0.43 | 0.51 |
| Year*BUSY*BIG4 | F(5,1242) = 1.78 | 0.11 |
| Year*BUSY*Year-end (related to COVID 19 extension) | F(5,1242) = 1.70 | 0.13 |
| Year*BUSY*TENURE | F(5,1242) = 1.27 | 0.28 |
| Year*BIG4*Year-end (related to COVID 19 extension) | F(5,1242) = 6.25 | 0.00 *** |
| Year*BIG4*TENURE | F(5,1242) = 2.41 | 0.04 ** |
| Year*Year-end (related to COVID 19 extension)*TENURE | F(5,1242) = 3.30 | 0.01 ** |
| BUSY*BIG4*Year-end (related to COVID 19 extension) | F(1,1242) = 1.78 | 0.18 |
| BUSY*BIG4*TENURE | F(1,1242) = 1.29 | 0.26 |
| BUSY*Year-end (related to COVID 19 extension)*TENURE | F(1,1242) = 1.95 | 0.16 |
| BIG4*Year-end (related to COVID 19 extension)*TENURE | F(1,1242) = 11.64 | 0.00 *** |
| Year*BUSY*BIG4*Year-end (related to COVID 19 extension) | F(5,1242) = 2.10 | 0.06 |
| Year*BUSY*BIG4*TENURE | F(5,1242) = 1.37 | 0.23 |
| Year*BUSY*Year-end (related to COVID 19 extension)*TENURE | F(5,1242) = 2.91 | 0.01 ** |
| Year*BIG4*Year-end (related to COVID 19 extension)*TENURE | F(5,1242) = 2.10 | 0.06 |
| BUSY*BIG4*Year-end (related to COVID 19 extension)*TENURE | F(1,1242) = 2.45 | 0.12 |
Note(s): ***, ** Indicate significance at the 1 and 5% level, respectively
Source(s): Authors' own creation
Figure A1
The line graph shows the interaction between year and COVID-19 extension status on the variable log 10 (C Y F E E). The title reads: “Interaction F(5,1242) equals 2.64, p equals 0.02,” and a note below states: “Vertical bars denote 0.95 confidence intervals. Letters apply to: Year-end (related to COVID-19 extension) only exclamation mark.” The horizontal axis is labeled “Year” and covers 2015 through 2020 with an increment of 1 year. Below it, the text reads, “Year-end (related to COVID-19 extension),” showing a legend where a blue square line represents category “0” and a red circle line represents category “1.” The vertical axis is labeled “log 10 (C Y F E E)” and ranges from about 3.6 to 3.9 with an interval of 0.1. The blue line (category 0) starts around 3.723 just before the year 2015, slightly declines until 2018, then increases sharply to approximately 3.899 just before the year 2020. Data points for 2015 through 2020 are marked with the letter “a,” indicating statistical grouping. Vertical error bars for each point show 95 percent confidence intervals. The red line (category 1) starts lower, around 3.625 just after the year 2015, and rises gradually each year, reaching about 3.771 just after the year 2020. Letters alternate between “b” in 2015 and 2019–2020 and “a” in other years, showing differences in statistical significance between points. The vertical error bars indicate variability for each year’s estimate. Note: All numerical data values are approximated.ANOVA means graph for the interaction of year and financial year-ends that resemble the COVID-19 financial reporting extension (2015–2020). Source: Authors' own creation
Figure A1
The line graph shows the interaction between year and COVID-19 extension status on the variable log 10 (C Y F E E). The title reads: “Interaction F(5,1242) equals 2.64, p equals 0.02,” and a note below states: “Vertical bars denote 0.95 confidence intervals. Letters apply to: Year-end (related to COVID-19 extension) only exclamation mark.” The horizontal axis is labeled “Year” and covers 2015 through 2020 with an increment of 1 year. Below it, the text reads, “Year-end (related to COVID-19 extension),” showing a legend where a blue square line represents category “0” and a red circle line represents category “1.” The vertical axis is labeled “log 10 (C Y F E E)” and ranges from about 3.6 to 3.9 with an interval of 0.1. The blue line (category 0) starts around 3.723 just before the year 2015, slightly declines until 2018, then increases sharply to approximately 3.899 just before the year 2020. Data points for 2015 through 2020 are marked with the letter “a,” indicating statistical grouping. Vertical error bars for each point show 95 percent confidence intervals. The red line (category 1) starts lower, around 3.625 just after the year 2015, and rises gradually each year, reaching about 3.771 just after the year 2020. Letters alternate between “b” in 2015 and 2019–2020 and “a” in other years, showing differences in statistical significance between points. The vertical error bars indicate variability for each year’s estimate. Note: All numerical data values are approximated.ANOVA means graph for the interaction of year and financial year-ends that resemble the COVID-19 financial reporting extension (2015–2020). Source: Authors' own creation
Figure A2
The title at the top reads: “Interaction F (5,1242) equals 6.25, p less than 0.01,” with a note below stating: “Vertical bars denote 0.95 confidence intervals. Letters apply to: Year-end (related to COVID-19 extension) only exclamation mark.” In both graphs, the horizontal axis is labeled “Year” and spans from 2015 to 2020 with an increment of 1 year. Below it, the text reads, “Year-end (related to COVID-19 extension),” showing a legend where a blue square line represents category “0” and a red circle line represents category “1.” The vertical axis is labeled “log 10 (C Y F E E)” and ranges from about 3.5 to 4.0 with an interval of 0.1. The left panel is labeled “BIG 4: 0.” In this panel, the blue line (category 0) starts at approximately 3.706 just before the year 2015, dips until 2018, and then sharply increases to around 3.924 just before the year 2020. Letters for blue points are “a” in all years, placed above the error bars. The red line (category 1) starts lower, near 3.59 just after the year 2015, rises slightly until 2017, dips in 2018, and then increases gradually to about 3.711 just after the year 2020. Letters for red points alternate between “b” in 2015, 2019, and 2020, and “a” in other years. Vertical error bars represent 95 percent confidence intervals for each data point. The right panel is labeled “BIG 4: 1.” In this panel, the blue line (category 0) begins around 3.743 just before the year 2015, remains fairly stable until 2018, and then increases steadily to about 3.876 just before the year 2020. Letters for blue points are “a” in all years. The red line (category 1) starts at about 3.664 just after the year 2015, rises slightly in 2016, dips in 2017, then increases sharply, reaching about 3.829 just after the year 2020. Letters for red points are “a” for most years except 2017, where it is “b.” Both panels show that category 0 generally trends higher than category 1. Note: All numerical data values are approximated.ANOVA means graph for the interaction of year, Big 4 and financial year-ends that resemble the COVID-19 financial reporting extension (2015–2020). Source: Authors' own creation
Figure A2
The title at the top reads: “Interaction F (5,1242) equals 6.25, p less than 0.01,” with a note below stating: “Vertical bars denote 0.95 confidence intervals. Letters apply to: Year-end (related to COVID-19 extension) only exclamation mark.” In both graphs, the horizontal axis is labeled “Year” and spans from 2015 to 2020 with an increment of 1 year. Below it, the text reads, “Year-end (related to COVID-19 extension),” showing a legend where a blue square line represents category “0” and a red circle line represents category “1.” The vertical axis is labeled “log 10 (C Y F E E)” and ranges from about 3.5 to 4.0 with an interval of 0.1. The left panel is labeled “BIG 4: 0.” In this panel, the blue line (category 0) starts at approximately 3.706 just before the year 2015, dips until 2018, and then sharply increases to around 3.924 just before the year 2020. Letters for blue points are “a” in all years, placed above the error bars. The red line (category 1) starts lower, near 3.59 just after the year 2015, rises slightly until 2017, dips in 2018, and then increases gradually to about 3.711 just after the year 2020. Letters for red points alternate between “b” in 2015, 2019, and 2020, and “a” in other years. Vertical error bars represent 95 percent confidence intervals for each data point. The right panel is labeled “BIG 4: 1.” In this panel, the blue line (category 0) begins around 3.743 just before the year 2015, remains fairly stable until 2018, and then increases steadily to about 3.876 just before the year 2020. Letters for blue points are “a” in all years. The red line (category 1) starts at about 3.664 just after the year 2015, rises slightly in 2016, dips in 2017, then increases sharply, reaching about 3.829 just after the year 2020. Letters for red points are “a” for most years except 2017, where it is “b.” Both panels show that category 0 generally trends higher than category 1. Note: All numerical data values are approximated.ANOVA means graph for the interaction of year, Big 4 and financial year-ends that resemble the COVID-19 financial reporting extension (2015–2020). Source: Authors' own creation