Both numbers are important, as "total union membership rising/falling" is useful to know if efforts to expand membership have been successful.
If a sudden rash of unemployment in an unorganized sector of the economy caused the total unionized percentage to surge, I wouldn't consider that a good thing. The ideological cause of unionization isn't inherently served by arbitrary fluctuations in the overall employment pool. Now, if Microsoft lays off 10% of its workforce and there's a sudden rush towards labor organizing as a response, sure... but I haven't seen that happening.
Both numbers are important, as "total union membership rising/falling" is useful to know if efforts to expand membership have been successful.
If a sudden rash of unemployment in an unorganized sector of the economy caused the total unionized percentage to surge, I wouldn't consider that a good thing. The ideological cause of unionization isn't inherently served by arbitrary fluctuations in the overall employment pool. Now, if Microsoft lays off 10% of its workforce and there's a sudden rush towards labor organizing as a response, sure... but I haven't seen that happening.