Sometimes we set the average TFP such that output is normalized to 1 in steady state (i.e. y=af(\cdot) s.t. y=1). Nonetheless in my model, the labor endowment is also normalized to 1 (i.e. unemp. + out\;of\;labor + labor=L=1), is that right? Thanks!!
Typically, you can assume both. The a is the necessary constant to allow scaling y while the utility function is usually written down in a way that requires a normalization of labor as well, e.g. \ln(1-n_t)