Use previous data to fill in blanks in table

Hi REKLP

I had a look at your file. I think the problem probably lies with using the fact tables to filter rather than the dimension table. That is dangerous and can easily bring about unpleasant side effects. I think we've been trying  unnecessarily intricate approaches while what you require can be done in a simpler manner.

Instead of the measure that you have:

m_ActiveEmployees =
CALCULATE (
    SUM ( ActiveEmployee[Active] );
    FILTER (
        ALLEXCEPT ( ActiveEmployee; region[ED_REGION] );
        ActiveEmployee[Year-Month] <= MAX ( ActiveEmployee[Year-Month] )
    )
)

try this one and, I believe, the issue will be no more. Same approach for terminated employees.

m_ActiveEmployees_V2 = CALCULATE (
    SUM( ActiveEmployee[Active] );
    FILTER (
        ALL(MonthTable[Year-Month]);
        MonthTable[Year-Month] <= MAX ( MonthTable[Year-Month] )
    )
)

Now there are two things to take into account with this approach:

1. Do you want to do the cumulative count per year only? If so, you'll to need to add the current year to the filter because as the  V2 is now, it accumulates guess since the beginning of time in your 'MonthTable table (1955)

2. If it's actually per year you'll have to decide what to do if the first month is blank. Right now it will be left blank. What would the previous value have to be in that case?

Let me know if this does the trick. Here is the file with the new measures.