Interpretation interaction term with a dummy variable in it  reg inc age educ men men_age
------------------------------------------------------------------------------
         inc |      Coef.   Std. Err.      t    P>|t|     [95% Conf. Interval]
-------------+----------------------------------------------------------------
         age |   .7339817   .0719376    10.20   0.000      .590875    .8770885
        educ |   .0536394   .0494509     1.08   0.281    -.0447341     .152013
         men |   -.690055   .4454689    -1.55   0.125    -1.576235    .1961245
     age_men |   .0888071   .0736333     1.21   0.231    -.0576729    .2352872
       _cons |   5.147985   .3095237    16.63   0.000     4.532244    5.763726

men is a dummy variable where men = 1, women = 0
Can I interpret the coefficient for the interaction as the following: A male worker earn $6 (-.69+.0888 = -.6) less than a female worker?
 A: Unless you have a specific reason to leave 'age' out, it's a good idea to keep the terms in your model that are involved in an interaction.  
Right now, your regression is fitting income as a function of two variables and a constant term.  You can understand the meaning of your interaction term by working slowly through the model piece by piece.  You could write your model like this:
$$ y = b_{cons} + b_{men} x_{men} + b_{men/age} x_{men} x_{age} $$
Say that one of your independent variables is a woman, aged 34, i.e. $x_{men} = 0$ and $x_{age} = 34$.  How does this change your model?
$$ y = b_{cons} + b_{men} (0) + b_{men/age} (0) (34) \\ y = b_{cons} $$
Since your IV's gender was a woman, age doesn't matter in your expectation of income (since you haven't included it in your regression).  Now lets pretend we have two more IVs, $[x_{men} = 1, x_{age} = 56]$, as well as $[x_{men} = 1, x_{age} = 21]$,
$$ y_{1,56} = b_{cons} + b_{men} + b_{men/age}(56) $$
$$ y_{1,21} = b_{cons} + b_{men} + b_{men/age}(21) $$ 
Notice that $b_{men}$ is the same for all your male IV's.  It is the "constant term for males".  $b_{cons}$ is the same for men or women, thus, $b_{men}$ tells you the strength and sign of the correlation between income and... being male.  
The magnitude and sign of $b_{men/age}$ tells you how age changes men's income (or if being male is correlated to how much you make at what age).  In your results you gave, $b_{men/age} \approx .08$.  This is a positive number meaning that in your data as age increases, the income for males increased at a rate of 0.08*b_{cons} per year.  For males aged 33, $y = b_{cons} + (0.08)(33)$, at 34, $y = b_{cons} + (0.08)(34)$.  You can see that the slope of this line is 0.08.
What you should ask next is, is this increase also true for women?  Does the gender of my DV's really matter?  If it is irrelevant for you to examine a correlation between age and income independently of gender, then you don't need to include a $b_{age}$ term in your regression.  However, you may find that including age makes your 0.08 get smaller and fall below p = 0.05.  Both men and women are likely to make more money as they get older and progress in their careers right?
