st: whether to include year dummies when standard errors are clustered by year?

[gjhxmu@sina.com]

I have posted related content, and sjsamuels gave me some important resource and suggestions.

What doubts me is whether to include year dummies when 
standard errors are clustered by year?

Take one model for example:
y=a0+a1x+a2-a14 year dummies.

If standard errors are clustered by year,
Should I type,
reg y x yeardummies, cluster(year)
or
reg y x ,cluster(year)?

Thank your for any help.


Sincerely,