The debate with Bradford De Long

1 commento (espandi tutti)

Disclaimer: I have not yet watched the video and was agnostic and uninterested about the topic (too much debate) before reading the seventh argument on the DeLong’s blog via Panunzi’s comment.

“My opponent did not have a single piece of data or of theory to support his claim that the "stimulus" package will be effective.”

It is not clear which qualities a “piece of research” should have to be convincing and able to support the suspicion that public spending multipliers might be positive. It is true instead that there are a variety of estimates out there. A nice survey of the empirical evidence is “Understanding  the effects of government Spending on Consuption” by Gali Lopez-Salido and Vallés, whose overview of the evidence tends to favour the predictions of the Keynesian model over those of the neoclassical one. Also the IMF World Economic Outlook of April 2004 chapter 2 presents an empirical survey that in synthesis says as follows “while there is a wide range of estimates of the size of this multiplier (0.1-1.6), partly reflecting a range of technical assumptions, typical results from both recent regression exercises and large macroeconomic models suggest that the impact of a fiscal expansion is significantly positive….larger for government spending increases than tax cuts; and it wears off over time”. The last 2008 WEO is much more diffuse on the topic but less optimistic, but still a good reading.

“The only piece of research he quoted, supporting the idea that there is a public spending multiplier larger than one … is a paper (Ramey's, see above) that focuses mostly on military expenditure. The latter …cannot, just cannot, be used to assess the economic effects of expanding Medicare and Medicaid or financing state's expenditure these days.”

Ramey and similar papers are not the only evidence out there, but it is the only methodology at the moment that responds to basically all criticisms of the endogeneity of fiscal policies which could undermine the estimates via VAR and alike. Moreover, if the effect on output and employment is positive with military expenditures (which is the least productive of all conceivable expenditures and, as a bonus, is not related to the economic cycle), than you can be sure that that is the lowest bound  of a multiplier. By the way, the Ramey’s paper is not so optimistic about the multiplier.

The seventh argument is like this (from Panunzi):

“I'm going to fight the stimulus … because we cannot afford a big fight over fiscal stimulus…it will distract the political system from the real and necessary task: fixing the banking system.”

That’s right.
My question: what if the fiscal stimulus works splendidly? The financial system is kept afloat, things become less pressing and everything is lost in the mist of time. Will they live happily ever after with a banking system still in a shambles? (and ineffective for years like in Japan). After all the fiscal stimulus may cover up rather than solving the issue.