The Obama administration’s proposed American Jobs Act is heavily weighted towards payroll tax cuts – more than half of the total cost is accounted for by cuts on either the employee or employer side. And it’s widely assumed that the direct spending provisions (about $100 billion in mostly infrastructure spending) have the least chance of making it out of the legislative process. Should this worry us?

At least a little. From an “old Keynesian” perspective, payroll tax cuts should work pretty well. Money in peoples’ pockets should probably boost their spending. There’s microeconomic work suggesting that people do respond pretty robustly to increased cash-on-hand and the macroeconomic multipliers tend to show that payroll tax cuts are pretty decent stimulus – far outpacing most other tax cuts, though falling well short of direct spending and transfer  payments.

However, from a “new Keynesian” perspective, payroll tax cuts may be not only less effective than advertised, but actually outright contractionary.

Gauti Eggersston has written a number of papers about how to think about the effects of macro policy when the economy is “at the zero bound” of short-term interest rates – like today’s American economy. In one paper he warns specifically about the contractionary possibility of payroll tax cuts. The (very) simplified intuition is that these tax cuts increase take-home wages and hence provide incentives to workers to increase the hours of work they supply the labor market. This increase in labor supply puts downward pressure on overall wages which pushes down prices. This price decline increases real interest rates (which are simply nominal rates minus inflation), which slow economic activity as well, thereby reducing aggregate demand.

As unemployment is simply the difference between labor supply and labor demand, payroll tax cuts exacerbate this gap on both sides – increasing labor supply and reducing economy-wide demand. During normal times, the Federal Reserve can short-circuit this vicious cycle simply by cutting nominal rates – but the short-term rates controlled by the Fed already sit at zero.

How seriously to take this warning?

On the one hand, John Taylor, an economist at Stanford and a vociferous critic of the Obama administration’s American Recovery and Reinvestment Act (ARRA), claims to have found no effect at all of temporary tax cuts on household consumption.

On the other hand, Taylor claims that even policies that are very well-targeted to cash-strapped households had no effect on spending – essentially arguing that even unemployment insurance and food stamp increases were saved dollar-for-dollar by recipient households. This seems implausible.

And, Dean Baker and David Rosnick re-examine Taylor’s results and show that temporary tax cuts do boost consumption by an amount greater than zero once one allows for a structural break post-2008 in the effect of stimulus on consumption (and, the rationale for assuming that the post-2008 economy is behaving very differently from what came before seems solid for pretty apparent reasons – Great Recession, Lehman Brothers, etc.).

Lastly, one should note that the payroll tax cuts of 2011 have not been accompanied by an obvious surge in labor supply – labor force participation rates fell by slightly more between January and August of 2011 than they did between the same months of 2010, even as the payroll tax cut should’ve induced more labor supply than the Making Work Pay tax credit it was frequently cited as “replacing”.

So where does this leave us on the question of payroll tax cuts? They probably do some good – even Eggersston notes that if they go to cash-strapped households that the “old Keynesian” effects may dominate the “new.” But it should be clear that a fiscal jobs bill weighted more than half towards payroll tax cuts is one clearly tailored at least as much for political traction as economic impact.

If the choice is “payroll tax cuts or nothing,” I’ll take the cuts. But, there are clearly more effective measures to spur job growth (the direct spending components of the American Jobs Act, for example) and the economy will be better off if these are part of any final legislation.