Global Journal of Management and Business Research, D: Accounting and Auditing, Volume 22 Issue 2

Table 4: Descriptive Statistics Variables Obs Mean Std. Dev. Min Max Ad_Lag1 1474 0.7112 2.3244 -13.7824 22.4344 Ad_Lag2 1474 0.2371 3.9847 -24.1736 25.2864 Tie_Lag1 1474 0.4260 0.7927 0.0000 7.7458 Tie_Lag2 1474 0.3936 0.6149 0.0000 6.8621 And 1474 0.8455 3.2638 -21.9746 29.1549 Size 1474 8.0550 1.0700 4.5356 11.9060 Tdta 1474 0.4796 2.2174 0.0000 0.9992 Dltrs 1474 0.1583 0.9260 -9.1430 10.1283 Roe 1474 0.1883 1.4189 -11.0637 11.1265 Roe_Lag1 1474 0.1687 1.7213 -17.4032 15.6130 Chexta 1474 1.7881 1.9720 0.0000 13.5859 Roa 1474 -0.0442 0.9224 -9.3040 7.0910 Roa_Lag1 1474 -0.0288 0.7803 -10.8105 5.8170 End_Lt 1474 0.1048 0.2450 0.0000 1.6370 End_Lt_Lag1 1474 0.1039 0.2489 0.0000 1.8609 End_Lt_Lag2 1474 0.0969 0.2366 0.0000 1.6425 End_Ct 1474 0.8347 0.8343 0.0000 6.1371 End_Ct_Lag1 1474 0.9241 1.0601 0.0000 9.6844 Table 5: Comparison of Discretionary Accruals by Firm Size b) Specification of the Empirical Model In this work, we use a simultaneous equation model to verify the existence of the link between accounting data management and taxation on the one hand, and to identify the exogenous financial ratios explaining these jointly endogenous variables on the other. The model is specified as follows: 1 = 21 2 + 31 3 + ⋯ + 1 + 11 1 + 1 K+ 1 2 = 12 1 + 32 3 + ⋯ + 2 + 12 1 + 2 K+ 2 Where ( i = 1, 2) are the endogenous variables, ( j = 1, ..., K ) are the exogenous variables. The following final model is estimated: 1 = 21 2 + 11 1 + ⋯ + 1 + 1 eq1 2 = 21 1 + 12 1 + ⋯ + 2 + 2 eq2 Where and are respectively the discretionary accruals ratio used as an earnings management measure, and the effective tax rate for firm t . The structure of the model used can be justified by the fact that the company manages its results according to the tax requirement, and aims to maintain a certain logic in the information transmitted in order not to draw attention of the tax auditors. It would therefore be inappropriate to think that certain exogenous factors that determine the accrual book values of a company are not in turn influenced by the latter. Thus, on the one hand, an upward (downward) change in the tax requirement may persuade managers to use their discretionary power in the financial information production process. On the other hand, managers’ with discretionary power may affect the effective level of taxation of their firms according to the objective they wish to achieve. Therefore, the model specified above is estimated using the two-stage least squares method which accounts for the possible endogeneity feature in the relationship between taxation and accounting data management. Taxation and Accounting Data Management: An Empirical Study in Cameroon 73 Global Journal of Management and Business Research Volume XXII Issue II Version I Year 2022 ( )D © 2022 Global Journals Ttest Ad , by ( pmege ) Unequa Two-sample t test with unequal variances Group Obs Mean Std. Err. Std. Dev. [95% Conf. Interval] 1 1327 -0.9405 0.0998 3.6340 -1.1362 -0.7448 2 147 -0.1488 0.0975 1.1819 -0.3415 0.0438 combined 1474 -0.8616 0.0905 3.4760 -1.0392 -0.6840 Diff -0.7917 -0.7917 -1.0657 -0.5178 diff = mean(1) - mean(2) t = -5.6765 Ho : diff = 0 Satterthwaite's degrees of freedom = 546.012 Ha : diff < 0 Ha: diff != 0 Ha: diff > 0 Pr(T < t) = 0.0000 Pr(|T| > |t|) = 0.0000 Pr(T > t) = 1.0000