Reading this post over at Free Exchange crystallized a few thoughts about stimulus that I've been having. First thing I agree with A.S. that "I am not convinced that cutting payroll taxes is so effective."
A.S. has some very good thoughts on some of these issues, for me however, I just feel there is a huge incidence problem. Payroll tax cuts do very little to make workers cheaper, the pay increase is pretty small so I don't see it having much impact on consumption, and a certain chunk of it will inevitably go towards paying down debt. The big problem though is that its impact is so diffuse and we have such high capacity underutilization. While I'm sure there is some impact on employment, it seems to me that a large chunk of this form of spending will be absorbed by slightly higher rates of utilization of existing employees and capacity. This isn't nothing, but I don't see it having much bang for the buck.
While I'm normally opposed to targeting, I think stimulus is one of the few policy areas where it pretty much has to be. A big part of downturns is psychological, since people's expectations are for low growth they invest and hire as if it were. There are two immediate logical implications of this (I admit upfront that I am sceptical of making arguments through logic without empirical backing, what backing I have is largely from the fairly weak impact of stimulus that has largely been diffuse). First, that stimulus spending has to be seen as being large enough to make a real difference. If the overall stimulus is fairly small, this means that it has to be carefully targeted to areas it can impact more strongly with the hopes that seeing an upturn in one area will have knock on effects in other areas. Second, the stimulus has to be sufficiently prolonged to impact expectations. I think this has been a big problem with much of the stimulus. While we want to make it temporary, we don't want to make it so temporary that a business can meet new demand with additional overtime or temp workers. Much of the last stimulus was aimed at lasting for a 2 year period, I think this may have been too short. For this reason, I tend to like direct spending, such as infrastructure spending, far more than I like fiddling with tax rates. Putting in place a 5 year bridge building project will create some medium term demand both for suppliers and for workers in the region it is being built, this is probably too long to ask people to do overtime for and long enough for new businesses to feel confident to step in to fill new demand. Two years isn't. The downside is the threat of white elephant projects but there's never such a thing as a purely optimal policy, the question is deciding which risks are most problematic.
I'll also add that I like the idea A.S. suggests about subsidizing part time workers, this also helps reduce the problems with disincentives I like to point out regarding unemployment benefits.
Though I have no idea what is meant by unpredictable government policies, unless the debt ceiling debate is meant. The only industry under considerable pressure is the financial industry but I don't see why this does much more than impact considerations regarding funding through debt vs. equity. I just don't see how this could be having a large real impact on businesses.