Management Literature and Its Implications for Standard Setting.

Table 5 shows that the rate of companies’ real earnings
management in the US is 0.1653 higher than companies’
real earnings management in Germany that is -0.5028.
The result of t-test showed that t value is 22.046 (p=0.00).
This finding indicates that the real earnings management
happened in the U.S. at 2004-2007 periods are different
from the companies in Germany. This difference is
concerning to the manipulation over real activity
performed by companies in the U.S. is higher than
manipulation over real activity in Germany companies.
Hence, the second hypothesis stated that real earnings
management of companies in the U.S. is higher than real
earnings management of companies in Germany is
supported at 1% level.
The result of second hypothesis probably indicate
that the tight accounting standards through rule-based
accounting standards approach used by the US causes
higher real earnings management practices as compared
to the level of earnings management exposures in
Germany, which use sprinciple-based accounting
standards approach. The result of second hypothesis
is consistent with results by Demski (2004), Ewert and
Wagenhofer (2005), Cohen et al. (2008) and Yu (2008).
The tests of each proxy of real earnings management
also demonstrate a consistent result with the test of
REM variable, thus, the results showed that t-test result
values are -11,461 for R_CFO, 3.563 for R_Prod, and
-7.686 for R_DEX with probability of 0.000. See table 6.
The test towards each proxies of real earnings
management showed a consistent result with the test of
REM proxy. The mean of R_CFO proxy in the US
(-0.2359) is lower than in Germany (0.0081). The
lower (negative) value of R_CFO indicates that the
higher companies conducted earnings management
through sales manipulation. The test of independentsamples t test by assumption of different population
variance showed the mean difference of R_CFO in The
U.S. and Germany is significant at the 1% level. This is
showing that the level of real earnings management
through sales manipulation in the U.S. is higher than
real earnings management via sales manipulation in
Germany.
The average mean for R_Prod proxy in the U.S. is
also higher than Germany that is -0.1300 for The U.S.
and -0.4173 for Germany. The higher (positive) value of
R_Prod indicates the higher companies conducted
earnings management through overproduction. The test
of independent-samples t test showed the mean
difference of R_Prod in the U.S. and Germany is
significant at the level of á=1%. This demonstrates that
the level of real earnings management through
overproduction in the the U.S. is higher than what
happened in Germany.
The average mean of R_DEX proxy in The U.S. is
also lower than Germany that is -0.0593 for the U.S. and
0.0936 for Germany. The lower (negative) value of
R_DEX indicates the higher companies conducted
earnings management by means of cutting off
discretionary expenses (SG&A and R&D). The test of
independent-samples t test(1-tailed) by assumption of
equal population variance showed the mean difference
R_DEX in the U.S. and Germany is significant at the 1%
level. This is showing that the level of real earnings
management by means of cutting off discretionary
expenses (SG&A and R&D) in the U.S. is higher than
in Ge