The article below the fold advances the thesis that in future debt ceiling negotiations Democrats must consider the debt ceiling as expendable, and default as a possibility, in order to prevent worse economic damage that would occur from assumed replays of the most recent debate.
In the recent debt ceiling debate, it was taken as understood by liberals and Democrats generally that the consequences of default would be disastrous.
"I think it would be a calamitous outcome. It would create a severe financial shock," Bernanke told the Senate Banking Committee during his second appearance before Congress this week. "Treasury securities are critical to the entire financial system … A default on those securities would throw the financial system potentially into chaos."
Likewise, Treasury Secretary Tim Geither, in the National Journal:
"A default would inflict catastrophic far-reaching damage on our nation's economy, significantly reducing growth and increasing unemployment," said Geithner in the letter to Bennet which was dated May 13. "Even a short-term default could cause irrevocable damage to the economy."
This was also adhered to by the Republican leadership:
At a closed-door meeting Friday morning, GOP leaders turned to their most trusted budget expert, Rep. Paul D. Ryan of Wisconsin, to explain to rank-and-file members what many others have come to understand: A fiscal meltdown could occur if Congress fails to raise the debt ceiling.
House Speaker John A. Boehner of Ohio underscored the point to dispel the notion that failure to allow more borrowing is an option.
"He said if we pass Aug. 2, it would be like 'Star Wars,'" said Rep. Scott DesJarlais, a freshman from Tennessee. "I don't think the people who are railing against raising the debt ceiling fully understand that."
While freshman Republicans from the Tea Party caucus felt less urgency about the matter:
Jenny Beth Martin, a co-founder of the Tea Party Patriots (the grassrootsier wing of the movement) agreed that default -- and the requisite end in government payments for the programs that go with it -- could hurt the thousands of tea party voters she represents. But she said her members are willing to take the hit.
"If it injures the people that we represent, but it's benefiting the whole country, that's what we need to be concerned about," she said.
Analyses of the debt debate largely agreed that this asymmetry in urgency allowed the Republicans, relying on the votes of the Tea Party caucus, to credibly threaten default that the Democrats found impermissible, extracting concessions from the President and the Democrats who control as much of Congress as they do. The situation was often referred to as hostage-taking, though a better model might be the Ultimatum Game ("yes or no, take what I offer or we both get the worst payout").
It remains our unchanged view that default would be disastrous. What has changed now is that Republicans have been successful in using the debt ceiling to extract concessions, and intend to do so in to the indefinite future. From the floor of the Senate, Minority Leader Mitch McConnell made this clear:
"never again will any president, from either party, be allowed to raise the debt ceiling without being held accountable for it by the American people and without having to engage in the kind of debate we've just come through."
The Ultimatum Game is typically played just once in social science experiments, because someone who expects the positions of the players to be reversed, or to be punished repeatedly by a "receiver" player, will rationally adjust their offered payouts, and the experimental interest is generally in the first offer, to study ideas of fairness in power asymmetries.
Some observers suggest that if positions are reversed in the future, a minority of Democrats can credibly make the same threat against a Republican President who will not abide a debt default. However, this diary takes the assumption that Democrats and Republicans will retain their relative estimates of the cost of default, in which case the Republicans will retain their "first player" position in the Ultimatum game. In that case, the only way to enforce greater fairness is to punish asymmetrical offers with a view to future plays. This diary proposes to examine, admittedly from the point of view of a non-economist with other academic training (mathematics), the relative costs of doing so.
The strategy is as follows. The diary attempts to estimate the effect of a default, based on projections from U.S. officials in the most recent debate and academic studies of previous sovereign defaults. The simplifying assumption is made that tax policy (or at least revenue) remains the same and so increased borrowing costs represent reduced spending on other priorities. The diary then attempts to estimate the effects of repeated debt negotiations that end similarly to the ones most recently played out, pricing their effects at estimated intervals of two years in to the indefinite future.
DEFAULT COSTS
Academic economists have studied the effects of default on nations' economic well-being. A relevant recent paper from 2006 is "The Elusive Cost of Sovereign Defaults", here:
In this paper, we compile a quarterly database on emerging market defaults and run panel growth regressions controlling for crisis variables, both annual (to check consistency with previous results reported in the literature) and quarterly. We include several leads and lags to ensure that the results are not driven by dating errors. We find that, when we look at quarterly data, growth rates in the post-default period are never significantly lower than in normal times. Moreover, the evidence indicates that, contrary to what is typically assumed, the output contractions often attributed to defaults actually precede them. Indeed, defaults mark the inflection point at which output reaches its minimum and starts to recover.
This should not be interpreted as proof that defaults in general do not matter. On the contrary, much in the way of a standard liquidity run, most of the financial distress that precedes the default decision may be due to its anticipation. However, our findings have distinct implications from a policy perspective. If defaults were costly a posteriori, the decision to default should weigh these costs against the fiscal effort needed to service the debt. However, once the default is anticipated (and its concomitant cost brought forward) by the market, the formal decision to stop servicing the debt entails no tradeoff and is therefore optimal (and even overdue).
Emphasis mine. As counterpoint, it must be pointed out that in the present case, the markets did not anticipate a U.S. default (see here and here for examples); they got nervous, but the probability as measured in investor behavior and surveys of experts was always considered to be low. This is likely to remain so in future debt ceiling debates, and hence it seems reasonable to take a more pessimistic view of the post-event costs of default.
To estimate this, I turn to CBO director Douglas Elmendorf, as quoted in an interview in the Christian Science Monitor:
"Even a small increase in the perceived risk of Treasury securities would be very expensive for the country. If Treasury rates moved up by just [0.1 percent] over the next decade, that would add $130 billion to interest payments over the decade."
The worst impact of default on spreads over the long term cited in the "Elusive Costs" paper, Ozler 1993, was about 20 basis points, or 0.2 percent, which would cost $260 billion in the first decade and grow with interest after that. In the letter of the chairman of the Treasury Borrowing Committee to Secretary Geithner, he anticipates that
The consequences of a ratings downgrade would be significant, with the potential for Treasury rates to rise by a full percentage point for each one-notch downgrade.
This would entail costs of some $150 billion per year over the next decade. That figure does not include the effects of a default on growth or GDP. The "Elusive Costs" paper of 2006 found that such effects were secondary and minimal, though their sample did not have major industrial nations to examine. The closest effect would be Russia, which grew strongly after default. Thus, this argument will take the high-end scenario listed above of the effect on borrowing costs of a single one-notch downgrade -- $150 billion per year -- and assume that growth effects from anything other than the forced reorganization of spending would be minimal. This assumption is certainly challengeable.
DEBT DEAL COSTS
The costs of the debt deal, set down on paper, are easier to quantify.
Promptly after passage of the debt ceiling deal, U.S. markets fell on the widespread agreement that the deal would have a negative effect on growth.
Last week brought the disconcerting news that the economy grew no faster than the population during the first six months of the year, in part because of spending cuts by state and local governments. Now the federal government is cutting, too.
"Unemployment will be higher than it would have been otherwise," Mohamed El-Erian, chief executive of the bond investment firm Pimco, said Sunday on ABC. "Growth will be lower than it would be otherwise. And inequality will be worse than it would be otherwise."
He added, "We have a very weak economy, so withdrawing more spending at this stage will make it even weaker."
The deal cuts $2.4 trillion (assume that some deal will be made by the "Super Congress" and pass the Congress) from government spending over the next decade, or an average of $240 billion per year. However, the cuts are backloaded, and one Congress cannot necessarily bind the actions of another. Over the next two years, we have the following assessment from The Atlantic:
Government is shrinking already. Total government contributed negative-2% to GDP growth in the last three quarters. In the next two years, this plan will cut government by another $70 billion. That's only 0.5% of a $14 trillion economy -- or 0.25% over two years. It doesn't sound like much. But in a fragile country that barely growing at 1%, even a fraction of total government spending can make the difference between growth and stasis.
So $70 billion in cuts, without new revenue, over the next two years can be firmly counted on. Our analysis will now assume that this will be repeated in future Congresses if the behaviors of the Democrats and Republicans remain similar into the indefinite future.
COMPARISON
Under our assumptions, any given round of debt ceiling negotiations includes the possibility of a default which is worse, in its effect on spending, than the effect of a single round of cuts. Repetition of similar cuts, however, immediately roughly equals and if continued certainly exceeds the impact of the consequences of default. Upon this analysis, we render the following conclusion:
Avoiding default was the right choice in this round of negotiations. Repetition cannot be permitted.
The next time such a situation arises, Democrats must make clear that default is preferable to confirming the precedent the Republicans are attempting to set. More sharply worded, while the Republicans hew to dogmatic economic slogans such as no new taxes, Democrats must take a cold-eyed view of comparative costs of the options available. Default is bad. Repeatedly harming the economy to avoid default is eventually worse.
This analysis is based strictly on economic consequences. One may argue, for example, that a civilized nation honors its debts. If it is legitimate to consider such principles as the sanctity of the United States debt, it is also legitimate to consider such principles as the government's duty to provide an environment of economic growth and security for its citizens, especially its poorest.
The analysis is also quick and dirty. I invite a more rigorous analysis of the comparative consequences by professional economists or other people better positioned to provide quantitative guidance on this question.