After how much time would the nonlinear effects of the subsidy be expected to be apparent?

What hypothetical effect would discontinuation of the subsidy have: while it's initially resulting in labor shortage reduction and then after time t?

Are counterfactuals helpful for this problem?