Univariate Analysis
Univariate Analysis
Panel A: Mann-Whitney Tests for Differences in Means between Subsamples
Ext. Audit
OSP
No
Column B Variable
Column A
Column A
(column A)
(column B) (column C)
Column B
Column C
Column C
ASSETS 16. 231 11.149 11.681 5.082 4.550 -0.532 RDSALES 0. 008 0.009 0.012 -0.001 -0.004* -0.003* TROUBLE 0. 073 0.076 0.037 -0.003 0.036** 0.039** RECINV 0. 262 0.222 0.241 0.040 0.021 -0.019 RISKIND 0. 058 0.063 0.068 -0.005 -0.010 -0.005
ASSTGROW 11. 099 10.793 6.965 0.306 4.134** 3.828** ROA
0. 033 0.039 0.035 -0.006 -0.002 0.004 FORSALES% 0. 114
0.030* 0.037* GOVRNK 4. 899 5.255 4.895 -0.356 0.004 0.360 ACE 0. 250 0.603 0.555 -0.355** -0.305** 0.048 AUTHORITY 0. 600
-0.233** -0.166** 0.067 AC independ
AC meets 0. 588 0.603 0.627 -0.015 -0.039 -0.024 AC expertise
0.006 Sample firms
Panel B: Chi-square tests for differences in ACE
Firms outsourcing to
Firms not outsourcing to
Totals ACE = 0
ACE Value
external auditor
external auditor
56 61 117 ACE = 1
19 83 102 Totals 75 144 2 χ statistic = 10.55**
Panel C: Chi-square tests for differences in ACE*AUTHORITY
Firms outsourcing to
Firms not outsourcing to
ACE*AUTHORITY
Totals ACE*AUTHORITY=0 69
external auditor
external auditor
81 150 ACE*AUTHORITY=1 6
63 69 Totals 75 144 2 χ statistic = 6.975**
Note: *, ** represent significance levels at .05 and .01, respectively. All variables are defined in Table 2, with the exception of AC independence, AC meeting frequency and AC expertise. AC independence is a dichotomous variable coded ‘1’ for audit committees comprised entirely of independent directors, ‘0’ otherwise. AC meets is a dichotomous variable coded ‘1’ for audit committees that meet at least four times annually, ‘0’ otherwise. AC expertise is a dichotomous variable coded ‘1’ for audit committees containing at least one director with financial expertise, ‘0’ else.
TABLE 5 Logistic Regression Results Binary Use External Auditor Outsourcing
OUTAUD=b 0 +b 1 LN(ASSETS) + b 2 RDSALES + b 3 TROUBLE + b 4 RECINV +
b 5 RISKIND + b 6 ASSTGROW + b 7 ROA + b 8 FORSALES% +
b 9 GOVRNK + b 10 ACE + b 11 AUTHORITY + b 12 ACE*AUTHORITY + ε
Independent Variable
INTERCEPT ? -2.437 3.367 LN (ASSETS)
-0.308 0.543 ACE*AUTHORITY
-0.622 7.885** Observations 219 Pseudo R-squared 0.115
*, ** - differences are significant at p-levels of less than 0.05 and 0.01. All variable definitions per Table 2.
TABLE 6 Tobit Regression Results
Outsourcing Activities to Current External Auditor
AUDMIX%*OUTAUD% =b 0 +b 1 LN(ASSETS) + b 2 RDSALES + b 3 TROUBLE +
b 4 RECINV + b 5 RISKIND + b 6 ASSTGROW + b 7 ROA +
b 8 FORSALES% + b 9 GOVRNK + b 10 ACE + b 11 AUTHORITY
+b 12 ACE*AUTHORITY + ε
Model 1
Model 2
Dependent Variable
Dependent Variable
AUDRTMIX%*OUTAUD%
AUDNRMIX%*OUTAUD%
T Independent Variable
Expected Parameter
INTERCEPT ? -2.251 -2.222 -5.667 -3.457 LN (ASSETS)
0.778 0.565 0.098 0.551 RDSALES
-2.347 -1.099 -2.054 -0.905 TROUBLE
-3.501 -1.774* -2.158 -1.115 FORSALES%
3.722 1.871* 2.221 1.831* GOVRNK
-5.198 -0.704 -4.552 -0.649 ACE
-1.599 -4.224** -0.774 -0.745 AUTHORITY
2.444 0.011 1.077 0.198 ACE*AUTHORITY
-0.429 -3.011** -0.816 -1.013 Observations 219 219 Pseudo R-squared
*, ** - differences are significant at two-sided (one-sided if in predicted direction) p-levels of less than 0.05 and 0.01. All variable definitions per Table 2.
TABLE 7 Tobit Regression Results
Outsourcing Activities to other Outside Service Provider
OSPMIX%*OUTAUD% = b 0 +b 1 LN(ASSETS) + b 2 RDSALES + b 3 TROUBLE +
b 4 RECINV + b 5 RISKIND + b 6 ASSTGROW + b 7 ROA +
b 8 FORSALES% + b 9 GOVRNK + b 10 ACE + b 11 AUTHORITY
+b 12 ACE*AUTHORITY + ε
Model 1
Model 2
Dependent Variable
Dependent Variable
OSPRTMIX%*OUTAUD% AUDNRMIX%*OUTAUD%
T Independent Variable
Expected Parameter
INTERCEPT ? -3.773 -2.222 -2.853 -0.888 LN (ASSETS)
-0.452 -1.005 2.724 0.845 RDSALES
-2.289 -2.007* -1.082 -2.777** TROUBLE
4.115 2.432** 3.271 2.682** RECINV
2.664 1.012 1.076 0.597 RISKIND
-0.344 -0.487 -0.898 -1.002 ASSTGROW
0.225 2.053* 0.847 2.428** ROA
-1.477 -0.772 -3.729 -0.903 FORSALES%
2.555 1.773* 2.007 1.745* GOVRNK
-0.119 -0.547 -0.295 -0.654 ACE
0.998 -0.568 -0.111 -0.815 AUTHORITY
-1.554 -0.089 -0.008 -0.221 ACE*AUTHORITY
-0.112 -1.012 -0.052 -0.678 Observations
219 219 Pseudo R-squared
*, ** - differences are significant at two-sided (one-sided if in predicted direction) p-levels of less than 0.05 and 0.01. All variable definitions per Table 2.
TABLE 8 OLS Regression Results
External Auditor Outsourcers Subsample