Fifth and last. Not relevant to the USA. Back in the day when US unions weren’t totally feeble, MacDonald and Solow wrote a brilliant paper on collective bargaining and tax based incomes policy.
Imagine a world in which firms must negotiation with unions (for example imagine Europe). The unions have two aims — they want high wages and they want high employment in the sector they represent. This means that a GM&UAW right to manage contract which specifies wages and working conditions and allows management to choose output, investment, and employment is Pareto inefficient. It makes sense to specify wages and the level of production (firms must produce more than the amount that maximizes profits given wages).
If unions have power (hah!) and behave optimally (ha!) representing workers in general and not just workers who have lots of seniority and job security (ha ha ha ha ha) then things are better, but not as good as they could be.
If we collectively really want high employment but don’t care about wages in sectors, then we want to give the rational firms and workers modified incentives. This might be because we really care a lot about the unemployed and don’t care so much how high incomes of the employed are. For MacDonald and Solow it is mostly because they think that higher wages mean higher prices so an equal increase in dollar wages in all sectors has no effect on real wages — that is that real wages are really relative wages — always and automatically.
In any case, the proposal is to reward increased employment and penalize increased wages. This changes the efficient choices for the union and firms. In theory it causes higher employment and lower inflation. In practice it is alleged to have worked on the rare occasions in which it was tried.
I ask why penalize increases in wages. The same effect occurs in the model if one penalizes wages (and rewards employment). The focus on the change of wages was natural back when macroeconomists were worried about inflation (paper published 1984). It follows from accepting the existing inter-industry wage structure. It follows from unions being powerful back then so a proposal which would generally punish unionized workers would not get support from Democrats in congress. That was long ago (kids believe me — I was alive back then — things were different — also they still are that way in Italy).
So in the model as written (and published in a top economics journal the AER) the income tax causes increased efficiency. The proposal is to tax income and subsidize employment (that is have an income tax and an EITC).
In theory this should work. In practice it works.