Harvard economist Ed Glaeser gives his recommendation:
But if America does embrace another stimulus round, we should limit the government’s role to being the big borrower rather than the big spender. Cutting payroll taxes for lower-income workers who have just been unemployed is an example of stimulus through borrowing, rather than spending. The government isn’t actually spending money on government services; it’s just borrowing the money and giving it to newly employed workers.
The most orthodox models, derived from the logic of David Ricardo, suggest that this kind of inter-temporal shuffling of taxes has limited downside risk, as long as the government isn’t at risk of default. Consumers can prepare for expected future taxes by saving today’s tax cuts. Moreover, reducing the payroll tax has the added advantage of increasing the incentives to work during a downturn.
The case for more complicated tax tweaks that affect other behavior is weaker….
The case for more government spending on tangible government products is most problematic.
While it is easy to get all misty-eyed about the Tennessee Valley Authority, public spending on roads or high-speed rail can be enormously wasteful. At the extreme, spending a billion dollars on a bridge to nowhere may temporarily increase employment and gross domestic product, but it does so by burning a billion dollars on something no one wants. Infrastructure is serious business, and it is impossible to spend quickly and wisely.
While wading in ignorance, it’s best to avoid the paths near the most dangerous depths.
There is little downside to giving a tax break to previously unemployed low-income workers. Those dollars are being given to people who value them. Building a bunch of unneeded highways, conversely, is a road to waste. The political and economic case for a second stimulus is strongest if that stimulus means a temporary tax reduction and weakest if the package is yet another increase in the size of the public sector.
Full post can be found here.