Wednesday, July 6, 2011

Debt Ceiling Negotiations and the Dangers of Maximalism

The key to human advancement is our ability to build on the achievements of the past. The people who are bringing us cures to diseases and exploring the outermost reaches of space and the fundamental nature of matter would be swinging clubs and digging for edible roots if they didn't have access to everything people have done before.

Similar development happens in all areas of human endeavor.  A modern chess Grandmaster with a time machine would make short work of past world champions, manufacturing experts have a structurally superior understanding of their processes (not just better machines) and modern negotiators are far more effective at getting good results than their predecessors.

Unfortunately, advances in a field are not always captured by people for whom that field isn't a core competence.  Government in particular seems vulnerable to falling behind the advances of the business world and our current debt limit negotiations are a potentially disastrous example.

I'm going to do a post soon exploring how the current negotiating strategies by both parties could prove effective but today I'm going to look at what I think is the more likely scenario -- that they are inefficient, likely to lead to missed opportunities and have a very real potential to do us great harm.

The basic strategy taken by the Republican party has been maximalism backed up by ultimatum.  They have demanded massive reductions in spending and declared as a matter of principle that absolutely no tax increases can be part of the deficit reduction package.  If their terms aren't met they will not authorize an increase in the debt ceiling, which presumably would cause the United States to default on their debt.

To illustrate the weaknesses of this approach I want to explore it using a simpler example -- a company that makes customized heavy equipment negotiating a deal with a customer.  Both sides have a general sense that the deal is worth about $10 million, and in fact the seller knows that the deal is attractive at any price of $8 million or more while the buyer knows it's attractive for them at any price of $12 million or less.

A maximalist approach would see the buyer offer $3 million for the equipment, while the seller insists that they need $20 million.  Then the dance begins, with each offering concessions until finally they (hopefully!) complete the deal somewhere within the ZOPA. Even if they do make a deal, their approach is costly in many ways:

1. It's inefficient.  Negotiations are often time-consuming affairs.  At best, a maximalist approach still involves a much longer dance of mutual concessions until the parties reach the ZOPA.

Think how many months, how many high-level meetings and how many hours of staff time have been spent on the debt ceiling negotiations so far, with both sides still largely sticking to their maximalist positions.  Think of the wasted energy at the Treasury department as they buy time for negotiations (we theoretically hit the debt ceiling a while back; the August 2 deadline is when Geitner's maneuvers can no longer prevent a default on some of our obligations).  Only the most anti-government zealot would see this as a good thing, and it's certainly not something we'd like to emulate in our own negotiations.

2. It erodes trust.  If you offer me $3 million and I counter with $20, how has the negotiation affected our relationship by the time we agree on $10 million?  An unreasonable demand -- especially when both parties know it's unreasonable -- is also inherently a dishonest one.  This is compounded when we back up our offers with statements like, "We'll lose money if we sell it for less than $15 million," or "I'd be violating my duty to shareholders if I paid you more than $7 million."

3. It's a reinforcing trap.  Suppose you're buying something from me and I claim I can't sell it for less than $100 but eventually agree to $40.  The next time you (or anyone familiar with that negotiation) deals with me, they're going to discount my initial claim by 60%.  If I decide I've had enough with maximalism and make a reasonable initial offer to sell for $50 I'm going to find myself facing a client who is unwilling to pay above $20.  Like a vendor in a market where extreme offers and heavy haggling are commonplace, I lose the option to negotiate more reasonably.

4. It destroys opportunities to create value.  Suppose you're the buyer in our heavy equipment example.  The factory that is going to use the equipment is ahead of schedule, so you really want to complete the deal in a timely fashion.  Every month that you spend in negotiations will cost you $2 million in lost profit.  What you don't know is that the seller could speed up manufacture of your equipment at a cost of $500,000.

For modern negotiators this is easy money.  They will quickly identify that accelerating delivery is worth more to the customer than it costs the vendor and adjust the contract accordingly.  A maximalist is much less likely to find opportunities like this both because they are wasting effort on unreasonable demands but also because less information tends to be shared when trust is lower.

Identifying value creating opportunities is certainly possible for maximalists -- it's just harder.  Going back to the debt ceiling negotiations, for example, I think it's extremely unlikely that there are absolutely no tradeoffs involving tax increases that both parties would consider preferable.  Republicans have virtually taken those possibilities off the table -- or at the very least, made it politically costly to enact them, which may amount to the same thing.

5. It can kill a good deal.  We like to assume that people are generally rational and that given the opportunity to make money or burn it we'll choose to make it every time. Reality does not always bear this out.  Negotiation history shows many examples where deals like our hypothetical heavy equipment contract failed because each side became trapped by their own statements or demands.

I think the odds strongly favor a deal being done to increase our debt ceiling but I don't for a moment consider it a foregone conclusion.  Democrats and Republicans have made very public statements about what they can possibly accept and those statements have no ZOPA.  As the rhetoric heats up, each party adds to the political cost of giving in to the other's demands and, from a psychological standpoint, increases the risk that they succumb to a number of cognitive biases.  As one observer pointed out, a game of chicken can end tragically even if both parties are willing to swerve; in Rebel Without a Cause, the decision to bail out was "trumped" by a jacket getting stuck to a car door.

In this case we're all in the car with them.  For my part, I wish we were playing a more constructive game than chicken.


  1. I think it's extremely unlikely that there are absolutely no tradeoffs involving tax increases that both parties would consider preferable.

    Both sides, but especially the Republicans, care more about how they are seen to have negotiated, than what the actual outcome is. Obviously, politicians, whose goal is to wield power and win elections, are going to behave very differently than business negotiators, whose goal is to make their business more successful (although they often also care about who gets the credit for that success). You seem to assume, in the above comment, that the objective function that each side is maximizing depends only on what the final result is, and not on how they got there. But in politics I think that is very, very wrong, because the negotiation is played out almost completely in public, and the opinion of that public is what is most at stake.

    I also think you underestimate the value of intransigence, or commitment. All you have to do is watch Dr Strangelove. If the Republicans can convince everyone that they are captive to the insane Tea Party types and can't possibly agree to any revenue increases, then they become much more likely to get a result with no revenue increases. Remember here that the ZOPA is very, very, very large, because the BATNA is so bad for everyone. So an objective party who cares more about the country than about politics (who wouldn't actually exist in these negotiations, but I digress) might agree to almost anything rather than accept default. When the ZOPA is so wide, the attractiveness of intransigence becomes very high. You give a business example with a very narrow ZOPA, which makes it not a very good analogy.

  2. Hi David,

    Your point about differing motives for politicians vs. businessmen is correct; politics can often be close to zero-sum in ways that business negotiations rarely are.

    The business example was chosen mainly for its simplicity, rather than as a perfect analogy. That said, I do think that politicians care about policy and that policy success is relevant to their long term political success. Moreover, there are certainly tradeoffs where incumbents from both parties benefit -- thus, even if we assume that politicians care only about electoral implications the tradeoffs I'm talking about should exist.

    Regarding intransigence/commitment, that's a big part of what I hinted at early in the post. The main way that the Republican strategy may succeed is precisely that by committing themselves to an extreme position they may force Democrats to make much bigger concessions to avoid disaster than if they negotiated "normally". They are like the person in the game of chicken who waves his steering wheel at the opponent, making it clear that he cannot swerve. Such tactics can work, but they are very high-stakes. Moreover, as you note, they are more likely to work with a counterpart who cares primarily about the national welfare. A counterpart who cares mainly about political gain has an incentive to meet intransigence with "reasoned" intransigence.

  3. I do not think that policy success is relevant to politicians' long-term political success. Maybe this would be true if we had more than two parties, because both parties could lose to a new political force if their joint policies are not successful. But in the actual US system, where we are only ever going to have the two parties we have now (most political scientists agree), the overall good bad that gets done is irrelevant, all that matters is who gets the credit (for whatever good happens) or the blame (for whatever bad happens).

  4. U.S. electoral politics is almost zero-sum from the perspective of the two main parties as a whole, but not from the perspective of individual politicians. Even if we assume (which I do not) that politicians are motivated almost entirely by electoral implications rather than genuine views on policy and public service, they must consider their local elections which can certainly be affected by policy success.

  5. I agree there is a slight tendency for incumbents to be re-elected more often when things are going better, and less often when things are going worse. However I think there are many reasons why this does little, if anything, to systematically encourage better policy choices. First, the effect mostly affects the most junior legislators, who have the least actual power or influence. Secondly, the effect is weak because it often takes a long time for policy decisions to actually affect outcomes, and it's hard to associate particular decisions with particular outcomes, or to assess counterfactuals (would the outcome be better or worse if different decisions had been made). Third, and most important, demagoguery is more powerful than logic, especially when most voters are not well informed. Remember that most voters right now oppose raising the debt limit, period. I think most political observers would say that if the debt limit talks end in a stalemate, there is a protracted financial crisis which leads to a second recession, etc., this is mainly bad for the President and therefore good for Republicans, even if those same Republicans create the problem with their intransigence. It might not play out that way, but it seems like the most likely outcome. So there is at least some temptation for Republicans to cause a worse outcome, which is better politically for them.

  6. Chad - I agree that maximalist position is the most destructive political force during negotiations and in my opinion is a political facade to garner public opinion for their negotiating position. Politicians obviously differ in that way than businesses, public opinion generally doesn't affect a company's negotiating package (will the public buy more of my product if I negotiate better?). The frustration lies in the artificial stance of the party simply to buy future votes. So much of the value lost is simply unnecesary as it is an attempt to affect the invividuals ability to stay in their job. That is a tremendous disservice to the American people, democracy as a whole and on this particular deal - the global economy.

    Happy Birthday ---