Doug Holtz-Eakin, pooh-poohing an extended payroll tax holiday that might provide needed stimulus (though admittedly a very ninetieth-best way of doing this) says that, as a matter of economics, it doesn't matter whether you do it on the employer or the employee side. As a matter of economics, this is incorrect.
Standard economic models of tax incidence show that, at equilibrium, it doesn't matter whether you change the tax hit on the employer side or the employee side, because that only affects legal incidence, without transforming the factors that determine economic incidence. But in the short run (i.e, in transition to the equilibrium), it unmistakably DOES matter which side you do it on.
Suppose wages are sticky in the short run, whether for psychological reasons reflecting nominal prices or simply because people have existing short-term contracts. For example, suppose I have a contract with my employer specifying that it will pay me $12 per hour through the end of the year. This is undoubtedly an incomplete contract - it doesn't provide, for example, that this is conditioned on assuming continuation of current payroll tax policy, such that we will agree to the same "true" after-tax wage no matter which side Congress might pick for a payroll tax holiday. If the payroll tax is cut on the employee side, I pocket the increased after-tax income. No renegotiation until the next go-round (where it would more likely occur if the holiday were ongoing or expected to recur) Likewise, if the payroll tax is suspended on the employer side, it pcokets the extra change.
True, supply and demand are going to be affected, but the point is that to a considerable degree all this takes time to unfold.
So while Doug is likely right about permanently changing the payroll tax on one side or the other (leaving aside details such as the fact that minimum wage laws don't adjust for payroll tax nominal incidence), he is clearly wrong about the economics of, say, another one-year payroll tax holiday that is enacted with little advance notice and great uncertainty in advance about whether it would in fact happen.
There's a distributional difference in who pockets the money, and also a stimulative difference since the workers may on average have higher marginal propensity to consume than the owners of the businesses that would pocket the transition gain the other way around.
To be sure, it's still a far-from-great stimulus choice, given that even the richest workers get the payroll tax holiday as to their wages up to the Social Security tax cap, plus the fact that high-earners get more than those earning less than the cap amount.
I also realize that it's not going to happen, as the Republicans have finally located a class of tax cuts that they don't like - those that would be potentially stimulative in the run-up to 2012.
But nonetheless, let's try to discuss it honestly without misapplying standard economics truisms (about long-term incidence) outside their proper realm of application.