In the signaling game of education, the informed party, the worker, moves first. By taking costly actions, he tries to convince the firm of his true ability.
In contrast, the screening game arises when the uninformed party moves first. Before the worker acquires education, the firm can specify the wage rate for each education level. To ‘screen’, the wage is set so that only the high ability worker is willing to work for high education level. But that education level would be too costly/painful for the low ability worker to acquire—the wage cannot offset the cost of attaining high education level.
Technically, screening models are easier than signaling models to analyze. The main challenge for signaling games is the multiplicity of equilibrium, which diminishes the predictive power. The commonly used solution concept is perfect Bayesian equilibrium. One can also strengthen/refine it with further restrictions on the belief system. One such refinement is intuitive criterion.
[KELOWNA, CANADA, 9/2015]