Stimulus Can Backfire: Fiscal Policy

Start with the consumption multiplier

A recent posting on The Hedged Economist commented: “The public has completely out thought the Keynesians. Increase the stimulus and the public saves more in order to pay the inevitable higher taxes. Furthermore, drive down interest rates and they’ll just hold cash.”

“Liquidity trap is what economists call it when people hoard their cash. In the discussion of stimulus back in September 2010, this blog argued the multiplier wouldn’t be linear. Well, it seems the wizards in charge may have succeeded in producing a negative aggregate multiplier, a result that should be darn near impossible.” (See: “Who Killed Stimulus as a Policy Option”)

Take a step back from the generalization and one’s view about rational expectations eliminates the need for a lot of other questionable assumptions (or substitute rational responses as an assumption). To illustrate, this blog will use two items. The first will be the topic of this posting. It is a recent (September 9, 2011) piece by Mark Zandi that Moody’s Analytics (a.k.a. Economy.com) makes available (“An Analysis of the Obama Jobs Plan”). It addresses fiscal policy.

The second is an opinion piece by Alan Blinder from the September, 28, 2011 WALL STREET JOURNAL (“Ben Bernanke Deserves a Break”). It will be the subject of the next posting. It addresses issues related to monetary policy.

Zandi and Blinder are two economists whose analyses and opinions are worth knowing. They also coauthored the only analysis of the initial stimulus efforts that this blog reviewed. That review was posted in September 2010 before the actual effectiveness of the stimulus was known. “PART 3, PART 4, and PART 5 of that review all dealt with their analysis.

The above links to the relevant parts of the review are live. They contained the first cautions about some of the methodological issues that led to a stimulus that accomplished much less than expected.

Obama’s first stimulus failed by any definition including his own, and it almost certainly fell well short of the result Blinder and Zandi anticipated. Yet, it might seem naive to advocate a rational expectations assumption given the considerable evidence on irrationality developed by behavioral economists. That, however, isn’t the right question. The more appropriate question for those who need point estimates of macro responses is: Is a rational response a better assumption than the greater fool assumption implied by responses that are assumed to be fixed? Neither is going to be correct.

Interestingly, in their analysis and forecast of the impact of the initial stimulus efforts, Zandi and Blinder acknowledge that multipliers vary based on economic circumstances. They focus on capacity utilization ignoring other factors that influence the multiplier.

More recently (September 9, 2011) Mark Zandi in “An Analysis of the Obama Jobs Plan” makes a very explicit reference to changes in behavioral responses:

“Confidence normally reflects economic conditions; it does not shape them. Consumer sentiment falls when unemployment, gasoline prices or inflation rises, but this has little impact on consumer spending. Yet at times, particularly during economic turning points, cause and effect can shift. Sentiment can be so harmed that businesses, consumers and investors freeze up, turning a gloomy outlook into a self-fulfilling prophecy. This is one of those times.”

“Consumers and businesses appear frozen in place. They are not yet pulling back—that would mean recession—but a loss of faith in the economy can quickly become self-fulfilling.”

That comes very close to acknowledging a shift toward rational expectations in the following sense. If one expects bad times, the response may be to batten down the hatches in ways that ensure bad times. The difference is his assumption that there is an irrational fear motivating the loss of confidence. It could just be that the response isn’t to an irrational fear. It may be a totally rational response to previous and probable policies.

To illustrate, consider this quote from an article by another economist, Gene Epstein (BARRON’S October 1, 2011, “Big Stimulus, Little Effect):

“It [fiscal stimulus] used to be called the fiscal gas pedal. If a recession strikes, you stomp down on the accelerator to help get the economy out of the ditch, pushing the federal budget into deficit. Or more realistically, because balanced budgets have become a rarity, you make sure that this year's deficit is noticeably larger than last year's.”

It’s perfectly rational to expect continuous deficits to result in higher taxes, and to cut spending in order to prepare to accommodate the tax burden.

When discussing a new stimulus, it’s possible to argue that the failure was then: this is now. It is also extremely hard to estimate how short of expectations the initial stimulus was. However, there is an even more fundamental problem with the “that was then, this is now” argument. If unsupported, it can be used either to justify or question any stimulus.

If, however, the involuntary accumulation of debt that the first stimulus imposed on the public inhibited the effectiveness of the stimulus at all, one would expect that negative impact to be greater now. The debt, the rate at which the debt is expanding, and the focus on the debt’s implications are all greater now. Most importantly, no one, not even an ideologue, can pretend that the debt can just keep growing, although some politicians think they can convince the public someone else will have to do the paying down of the debt.

It’s important to remember the phrase “for those who need point estimates of macro responses” in the discussion above. There is a more practical approach. The reality is the “no change, stable multipliers assumption” and “the rational expectation, complete adjustment assumption” are both logical constructs that allow point estimates. However, they don’t even bound the possible. People could over-adjust to the debt and more than offset the stimulus. The practical approach is to view the potential responses as providing nothing more than a way to estimate the potential impact of stimulus.

What’s particularly dangerous about the current stimulus proposal is that it’s being proposed as being offset by taxes. If that were the case, the simulative impact is totally dependent upon a very questionable, small difference. It depends upon the positive impact on those being subsidized being greater than both the proposal’s direct negative impact on those providing the subsidy and any negative adjustment it induces among those negatively affected. While advocates of stimulus point to differences in timing as justification, it’s a weak defense in an environment where the timeframe just happens to coincide with the election cycle.

Further, it’s an election cycle that is focused on the issue central to the entire concept of government stimulus. As stated in PART 4 of the discussion of the initial stimulus efforts: “There are more basic questions about potential alternative policies. Basically, the underlying philosophy of the entire policy response should be questioned. It is not the need for a policy response, but the assumption inherent in parts of TARP and most of the fiscal stimulus that is questionable. Both are predicated on the assumption a trickle-down approach is best. In essence: give the money to a government, an investor, an automaker, etc., and just count on it to flow to the general benefit of the population.”

Current circumstances weaken Zandi’s confidence argument. If, as hypothesized above, there is a substantial portion of the population questioning the underlying philosophy of government stimulus, stimulus might actually undermine confidence. Further, the confidence issue is aggravated by differences in the resources (capital and incomes) of those who believe stimulus will help and those whose confidence would be undermined by another stimulus.

Some might consider it partisan to point out the high potential of failure of a new stimulus. The subtitle and focus of “Stimulus more or less? A failure not being acknowledged. PART 1,” was “The economic impact of fiscal stimulus isn’t a partisan issue. Nor should partisan leanings be the criteria for defining stimulus.” Beyond recommending that posting there isn’t much one can say.

Others may conclude it reflects an ideological predisposition. That’s nonsense. If people could over-adjust as individuals undermining stimulus, they are equally likely to over-adjust collectively through their democratic process. If they are the greater fool a constant multiplier implies (consumers that don’t respond to debt are foolish), then they would be equally likely to display their foolishness collectively through policy as individually as consumers. Ideological predisposition (and The Hedged Economist has his own) do not determine when stimulus is advisable or what impact it will have.