GuiAll
commented
3 months ago Reworking the tutorials, the solution would actually to compute the wage as the marginal value of labor within the hetblock itself, as part of the backward step, thus obtaining an "firm specific wage" (if the firm was a representative agent) w. Then the block would aggregate it as a weighted average wage W, which correspond to the wage paid on the labor market.
Or am I missing something ?
Greetings
I am building a model with heterogeneous firms with discrete adjustment costs, a la Khan & Thomas (2008).
I am struggling on the way to calculate wages, which require the distribution of firms over their capital stock and their idiosyncratic productivity.
First idea was to create a seperate function out of the blocks that would take the distribution from the firms Het block, calculate the wages over the capital and productivity grids with the distribution, and obtain the wages that would be feed into the relevant simple blocks (households and market clearing). But I am not sure it is legit, as, (from the tutorials), blocks seem to receive as inputs only outputs from other blocks and parameters.
Second idea was to create a simple block "labor market" that would determine the wages. But I understand from the RBC tutorial that only univariate variables can be input in simple blocks. However, the distribution is defined over two variables (productivity and capital), unless I misunderstand the "univariate" aspect ?
Thus is there a way to build it in the SSJ framework ?
Addedum : I see that hetblocks generate
DandDbegdistribution (Krusel & Smith and Tutorial 1 workbooks). What is the difference between the two?Thank you in advance for any help on the matter !