Tim Harford on Thomas Schelling

Florida and the War Chest

Nate Silver offers the following logic: winning Florida is essentially for Romney but not Obama; Obama has more paths to victory than Romney; Romney is currently winning in Florida; therefore Obama should concede Florida and re-direct his attention (and campaign spending) elsewhere.

If winning Florida represents a high-upside case for Mr. Obama, however, it also comes at considerable expense. 

Florida, because of its large population, is an expensive state to advertise in. And it is a state that probably does require a considerable advertising expenditure. Florida’s population is large, but not especially dense, spread out in a number of exurban and suburban communities throughout the state. It’s easier to reach voters through the airwaves there than by knocking on doors or appearing at campaign rallies. ... 

All of this should call into question whether Florida represents a wise use of resources for Mr. Obama. ...

Mr. Romney certainly doesn’t need Pennsylvania to win the election, but going for broke there is arguably a better strategy for him than having to pick off 4 or 5 states that are now tied or where Mr. Obama holds a small lead.

Mr. Obama, conversely, just needs to hold is ground in those same states. Trying to pull Florida back into his column would represent a heavier lift — and probably an inferior strategy given the recent polls there.

I'm not sure this is correct. Perhaps Obama should focus more on Florida because he's losing there, and because he has so many other ways to win the election. This may be true even if Obama expects that there is a pretty high probability he will lose Florida.

Why? Because Romney has to win Florida and Obama doesn't. If Obama dedicates a bunch of resources to Florida he can force Romney to do so the same to counteract Obama's move. Any money and time Romney is spending in Florida is money and time he isn't spending in Virginia, Ohio, Iowa, New Hampshire, Colorado, Nevada, Wisconsin, and Pennsylvania. Because Obama is winning in most of those states, and only needs to carry some of them to win the general election, anything he can do to prevent Romney from gaining in states other than Florida gives Obama an edge in the general election. Forcing Romney to expend resources defending his lead in Florida is one way to do that.

This only makes sense if Obama can credibly threaten Romney in Florida of course. But he probably can. Romney's lead in Florida is only a few points (Silver mentions a spread between 1-5% depending on the poll, so let's say 3%). If Romney loses Florida the election's all but over, so Obama would need a relatively low probability of victory there to force Romney to expend a lot of resources to guarantee a Florida win. Obama should force him to do it.

If Obama concedes Florida he's inadvertently putting other states in play by allowing Romney to divert resources from defending Florida to attacking Obama's advantage elsewhere. Obama should do whatever he can to prevent that from happening. Fighting in Florida means not fighting in Ohio, Wisconsin, and Pennsylvania. That's good for Obama, as he currently leads in all those states and will win the election if gets them all.

Those who have an education in game theory should recognize this dynamic as similar in some ways to a more complicated version of the "War Chest" game. The gist is that Obama should not expend his resources try to win the most states. He has an advantage: he's winning the election right now, and he gets no advantage from winning with 300 electoral college votes rather than 275. Instead, Obama should focus his efforts on preventing Romney from making further gains in states Obama currently leads. That involves strategic considerations that go beyond allocating resources to states where Obama has the best chance of winning the state. Instead, Obama should allocate resources to states where he has the best chance of winning the whole election.

I Can't Believe I'm Defending Waltz

Over at the Duck of Minerva, Patrick Porter takes issue with Kenneth Waltz's argument that it would be no horrible thing if Iran got the Bomb. Read those before you read this...

... So Porter's right about the trembling hand. But I don't think he's being fair to Waltz (and I'm not generally inclined to give Waltz the charitable reading, but the old man has a few points that Porter hasn't dealt with).

1. We're in probabilistic territory here. The question is not "would a nuclear exchange be horrible" or even "would the pursuit of more advanced weapons exacerbate some already-existing tensions". The question is "is the likelihood of bad outcomes higher or lower if Iran gets the bomb, relative to other feasible counterfactuals". In other words, we need to specify values for two variables: the probabilities for nuclear conflict and conventional conflict, and the likely cost of those conflicts. This is hard, but it's really the whole exercise. Suppose, over the next two decades, that there was a 1% chance of 10 million deaths from a nuclear exchange involving Iran, and a 30% chance of 1 million deaths from a conventional war involving Iran. If we had to choose between the two, we'd have to accept the possibility of a nuclear exchange as being objectively better. We don't get to choose "none of the above"... that's Waltz's point. Maybe you'd assign different probabilities and/or different costs to those outcomes, but the point is that Waltz is forcing us to be precise. It's not enough to just say "nuclear war would be horrible".

That's the whole reason why Waltz keeps citing India/Pakistan. Porter writes: "In the wake of 9/11, a Pakistani army general warned India that his country could launch a rapid nuclear attack, telling Alastair Campbell to remind the Indians: ‘It takes us eight seconds to get the missiles over.’ If this volatile frontier is a signpost of things to come in the Gulf, then the future is dark."

How so? It's been a decade since then, with no significant casualties. Tensions now are lower than they were before both states went nuclear, and the trend is in a positive direction. How is that a "dark future" relevant to feasible alternative scenarios?

2. Regarding missing Armageddon through luck... yes. And I agree that the trembling hand is not to be taken lightly. But it is not only present in nuclear exchange. Nuclear conflicts may be more likely to escalate more quickly than conventional wars -- although we don't really know, since it hasn't ever happened -- but other conflicts still escalate. Waltz's argument, which must be taken seriously, is that the very severity of a nuclear exchange mitigates the effect of the trembling hand. To use Porter's examples of close calls, the Soviet officers worked hard to dissuade their commander from firing nukes just in case it was a false alarm. The Kennedy administration held back in 1962 just in case there might be nuclear retaliation. They did these things precisely because nuclear war would be so horrible that they didn't want to chance it. At every knife-edge point we've had someone has made the decision that a nuclear holocaust would not be initiated by him. At every chance, the trembling hand has pulled back from the button. That observation does not mean that we'll always be so lucky, but it does imply that luck isn't the only factor in operation.

3. Suppose you disagree with my #1, and think that the probability of a nuclear exchange and/or the costs of one are higher than I've put down. If that is the case, then aren't you required to advocate for a pre-emptive strike against Iran? One can't have things both ways... either the risk of a nuclear Iran is acceptable or it is not. Porter doesn't make it clear where he stands on this question. But to criticize Waltz -- who does make it clear where he stands -- he must. Porter leaves the impression that Waltz is being flippant, but I don't think that he is. Waltz is making calculations. They may be "ahistorical" -- in fact they must be, since there is no history of mutual nuclear exchange -- and they may be based on Waltz's gut, but they are calculations nonetheless.

Waltz is right about one thing: there is a correlation between nuclear capability and peace. Perhaps the correlation is spurious. Perhaps not. But it is correlation, and it can't just be dismissed.

New Research

Are these results surprising? Not to me. And given the high barriers to entry for bargaining, we should expect to see regulations benefit large firms with a history of lobbying activity.

The Dynamics of Firm Lobbying
William R. Kerr, William F. Lincoln, Prachi Mishra 

NBER Working Paper No. 17577 

We study the determinants of the dynamics of firm lobbying behavior using a panel data set covering 1998-2006. Our data exhibit three striking facts: (i) few firms lobby, (ii) lobbying status is strongly associated with firm size, and (iii) lobbying status is highly persistent over time. Estimating a model of a firm's decision to engage in lobbying, we find significant evidence that up-front costs associated with entering the political process help explain all three facts. We then exploit a natural experiment in the expiration in legislation surrounding the H-1B visa cap for high-skilled immigrant workers to study how these costs affect firms' responses to policy changes. We find that companies primarily adjusted on the intensive margin: the firms that began to lobby for immigration were those who were sensitive to H-1B policy changes and who were already advocating for other issues, rather than firms that became involved in lobbying anew. For a firm already lobbying, the response is determined by the importance of the issue to the firm's business rather than the scale of the firm's prior lobbying efforts. These results support the existence of significant barriers to entry in the lobbying process.

This next one seems very inventive, in a "create your own science" kind of way. Has anyone else done anything like it?

Testing the Global Financial Transparency Regime
J. C. Sharman 

International Studies Quarterly Vol. 55 No. 4 

How can we tell whether rules that apply in theory actually do so in practice? Realists argue that the gap between what formal rules proscribe and their effectiveness may be particularly wide at the international level. Furthermore, dominant states may impose costly standards on others that they themselves choose not to implement. To test these propositions, the article assesses the effectiveness of international soft law standards prohibiting anonymous participation in the global financial system by seeking to break these standards. The findings indicate that the prohibition on anonymous corporations is relatively ineffective and is flouted much more in G7 countries than in tax havens. The article contributes to and extends the work of realist scholars in international political economy, both in their skepticism of formal rules and focus on the effects of power. Evidence is drawn from the author’s solicitations and purchases of anonymous shell companies from 45 corporate service providers in 22 countries.

The IPE work on exchange rate regimes continues to improve.

Fear of Floating and de Facto Exchange Rate Pegs with Multiple Key Currencies Thomas Plümper and Eric Neumayer 

International Studies Quarterly Vol. 55 No. 4

This paper adopts and develops the “fear of floating” theory to explain the decision to implement a de facto peg, the choice of anchor currency among multiple key currencies, and the role of central bank independence for these choices. We argue that since exchange rate depreciations are passed-through into higher prices of imported goods, avoiding the import of inflation provides an important motive to de facto peg the exchange rate in import-dependent countries. This study shows that the choice of anchor currency is determined by the degree of dependence of the potentially pegging country on imports from the key currency country and on imports from the key currency area, consisting of all countries which have already pegged to this key currency. The fear of floating approach also predicts that countries with more independent central banks are more likely to de facto peg their exchange rate since independent central banks are more averse to inflation than governments and can de facto peg a country’s exchange rate independently of the government.

And, lastly, an extension of Kydd's 2003 by UNC's Mark Crescenzi and co-authors:

A Supply Side Theory of Mediation
Mark J.C. Crescenzi, Kelly M. Kadera, Sara McLaughlin Mitchell, and Clayton L. Thyne 

International Studies Quarterly Vol. 55 No. 4 

We develop and test a theory of the supply side of third-party conflict management. Building on Kydd’s (2003) model of mediation, which shows that bias enhances mediator credibility, we offer three complementary mechanisms that may enable mediator credibility. First, democratic mediators face costs for deception in the conflict management process. Second, a vibrant global democratic community supports the norms of unbiased and nonviolent conflict management, again increasing the costs of deception for potential mediators. Third, as disputants’ ties to international organizations increase, the mediator’s costs for dishonesty in the conflict management process rise because these institutions provide more frequent and accurate information about the disputants’ capabilities and resolve. These factors, along with sources of bias, increase the availability of credible mediators and their efforts to manage interstate conflicts. Empirical analyses of data on contentious issues from 1816 to 2001 lend mixed support for our arguments. Third-party conflict management occurs more frequently and is more successful if a potential mediator is a democracy, as the average global democracy level increases, and as the disputants’ number of shared International Organization (IO) memberships rises. We also find that powerful states serve as mediators more often and are typically successful. Other factors such as trade ties, alliances, issue salience, and distance influence decisions to mediate and mediation success. Taken together, our study provides evidence in support of Kydd’s bias argument while offering several mechanisms for unbiased mediators to become credible and successful mediators.

What Do Regulations Do?

I believe I wrote about this year-old post by John Hempton awhile ago, but it's worth revisiting. I like the way he thinks about the effect of financial regulation:

In the UK banks were allowed to lever themselves to a silly extent (similar over-leverage occurs in their life insurance companies). Overleverage as a policy was the defining character of Northern Rock. 

Individually it makes sense for banks to lever up. However competition was intense - and collectively it was insane. Northern Rock was levered 60 times or so - but to mortgages that were really thin margin. Their spreads were about 40bps.   

What I suspect is happening is all the banks are standing on tippy-toes. It is individually rational - collectively insane because competition kills the benefit of all that extra leverage. Margins in the UK - the most over-levered market on the planet - fell further than anywhere else.  

Of course competition was good for borrowers - at least for a while. Lower spreads meant cheaper finance - but not dramatically cheaper. Spreads of 150bps on mortgages levered 15 times is about as profitable as spreads of 40bps levered 60 times. Competition might drive spreads down by 110bps - at the risk to the whole banking system.

In most of the popular discourse and academic literature banks are presumed to be opposed to regulation, because it corrects market inefficiencies by forcing firms to internalize negative externalities*. The public choice school** argues that there are times when this is not the case -- when incumbent firms can use regulatory structures to collect rents -- so there is no reason to start from the assumption that regulations will be welfare-enhancing.

Hempton is proposing something else: thinking of financial markets as creating a prisoner's dilemma for banks. In this case, banks would be better off if they were able to collude. They'd be able to maintain fairly large spreads, and thus profits, without taking on inordinate risk. The "cooperative" outcome is Pareto-optimal (for the banks at least). But it isn't individually rational. If all the other banks are maintaining higher standards, you can capture quite a lot of market share by "defecting" -- levering up, in this example. To do this you will have to accept lower margins, but profits will still increase if you increase volume enough.

Of course what is individually rational for one firm is individually rational for all firms, so just like in a prisoner's dilemma everyone "defects", driving down margins without capturing any more market share. In this scenario, bankers would prefer regulations like minimum capital adequacy and limits on leverage,  not because it bestows rents (at least not only for that reason), but because it changes the structure of the strategic interaction. Firms can now attain Pareto-improving outcomes where they can achieve a decent profit at fatter margins without taking on so much risk. So, ceteris paribus, in this situation firms should actually prefer to be regulated so long as everyone else is regulated too.

And, in fact, in the wake of the financial crisis every banker said they supported re-regulation of the financial sector so long as it affected everybody. But here's the kicker: the same dynamic that makes regulation Pareto-improving also makes regulatory avoidance very lucrative. If you can figure out a way to arbitrage the regulation, you can capture more market share at a slightly lower margin, thus boosting profits. In terms of the prisoner's dilemma, you can profit by defecting while everyone else cooperates. The rise of the shadow banking system is best understood in this light.

Meanwhile, I'm not as perplexed as Drezner is by recent developments in domestic and international regulatory regimes. First, Basel III went basically the way previous rounds went. This process isn't as simple as "bank preferences are communicated to national governments, and also includes the preferences of voters and policy elites. Second, the majority of the Dodd-Frank rules haven't been written yet, must less implemented so it's far too soon to say that bankers have "lost" in any meaningful way. Third, there are very real concerns that the EU won't be able to begin enforcing Basel III any time soon, which could potentially affect the competitiveness of US banks (this is what Jamie Dimon was talking about when he called Basel III "anti-American"). Fourth, Dodd-Frank contains dozens of provisions on top of Basel III, some of which could effect international competitiveness (although most won't).

Lastly, I think Drezner is making too much of the fact that the banks aren't getting everything they want. They are still getting quite a lot -- Dodd-Frank implementation is slow, underfunded, and every GOP candidate vows to repeal or otherwise castrate it -- but they never get everything they want. Finance is one of the most heavily-regulated industries in the country. They routinely lose political fights. The influence of bankers on politics is real, but it is quite often over-stated.

*A line of thought that began with Pigou, who wrote more about taxation than regulation, but the fundamental principle is the same.

**Notably Stigler and Peltzman, extended onto the international level by Oatley and Nabors.

New Research

On the Network Topology of Variance Decompositions: Measuring the Connectedness of Financial Firms Francis X. Diebold, Kamil Yilmaz
NBER Working Paper No. 17490
We propose several connectedness measures built from pieces of variance decompositions, and we argue that they provide natural and insightful measures of connectedness among financial asset returns and volatilities. We also show that variance decompositions define weighted, directed networks, so that our connectedness measures are intimately-related to key measures of connectedness used in the network literature. Building on these insights, we track both average and daily time-varying connectedness of major U.S. financial institutions' stock return volatilities in recent years, including during the financial crisis of 2007-2008.

This is important work, and I know that several regulators and central banks (including the Bank of England) are starting to take this sort of modeling -- weighted, directed networks -- very seriously. When you're trying to track sources of systemic weakness you really need to know what the system looks like. The problem isn't just "too big too fail", it's also about which firms are tightly connected to many other firms. These two will often correlate, but not always and not perfectly, so knowing the difference is important.

The Stock Market Crash of 2008 Caused the Great Recession: Theory and Evidence Roger Farmer
NBER Working Paper No. 17479
This paper argues that the stock market crash of 2008, triggered by a collapse in house prices, caused the Great Recession. The paper has three parts. First, it provides evidence of a high correlation between the value of the stock market and the unemployment rate in U.S. data since 1929. Second, it compares a new model of the economy developed in recent papers and books by Farmer, with a classical model and with a textbook Keynesian approach. Third, it provides evidence that fiscal stimulus will not permanently restore full employment. In Farmer's model, as in the Keynesian model, employment is demand determined. But aggregate demand depends on wealth, not on income.

I think some of this gets to my confusion about Keynesianism from a few days back. I think the last sentence particularly drives at what I was saying before: if the monetary multiplier is low because of expectations, then how can the fiscal multiplier be high under the same set of expectations? It makes more sense (to me) for behavior to be conditioned by wealth more than income, particularly if the income is temporary. I clearly need to become more familiar with Farmer's work.

And here's a near-complete preprint of Herb Gintis' most recent book, The Bounds of Reason: Game Theory and the Unification of the Behavioral Sciences. Via one of Phil Arena's commenters.

Schelling and the Euro

Ryan Avent:

It's really a mess. But one thing should be clear: it's in the interest of all the negotiating parties to be as apocalyptic in their warnings as possible. If the Greeks don't draw a hard line, they get a raw deal, and the same goes for the European Union, and the constituent governments, and the ECB, and the IMF. Ultimately, everyone expects that the negotiators will back down and an agreement will be reached, but in the mean time it's worth it to negotiate like a madman. The big downsides to this are, first, that it gives everyone reading newspapers a fright. And second, when so many parties are playing this brinkmanship game, there's always a risk that something goes awry and a deal isn't reached. Frankly, Europe isn't giving markets a lot of reason to be confident that the process of handling Greek insolvency is actually, underneath all the posturing, under control.

The logic of brinksmanship was put famously (and well) by Thomas Schelling. He described nuclear deterrence as two countries, chained together, dancing closer and closer to the edge a cliff, over which they'd both tumble if one of them slipped. And I've written similar things along those lines over the past year or so. At this point, however, I wonder how well the metaphor fits. The Greek political economy does not remotely resemble a unitary actor, at this point. Tens of thousands rioting in the streets, government turnover and possible recall, etc. It looks like Greece is racing full-steam towards the edge, attached to the rest of the EMU by a fairly thin cord, while Germany and France stand well back from the edge, holding scissors and debating whether and when to cut loose.

It's unclear whether this puts "Greece" at any bargaining advantage, because it's unclear what "Greece" represents and therefore what it desires. More accurately, there is an internal battle in Greece over the value of Euro membership. Germany and France can alter the terms of that membership on the margins, mostly by buying time, but they can't change the game. Even if some elites wanted to, Germany and France are limited by their domestic polities as well.

We need a new metaphor.

Misc.

Things have been busy around here lately, so in lieu of a substantive post here's some links:

-- Dani Rodrik vs. Turkey. This has been going on for awhile, but in light of the "democratic" revolutions spreading through the Middle East it's worth noting that democracy is a long, uneven process. Rodrik accuses some in Turkey of operating a "state within the state", that is anything but democratic. I'm in no position to evaluate his claims, but if true they are damning.

-- The paradoxical politics of credible commitment. The conclusion is "Neat economic theory trips over messy political reality once again", and it illustrates the importance, but difficulty, of accurately matching interests to actions. Among other things.

-- Paul Krugman's posting template, revealed. Brilliant.

-- Iraq's economy is not doing well.

-- I'm not sure why Stephen Walt is still so concerned with debates that everyone else abandoned two decades ago, but he is. Here he tries to stack the deck in favor of realism in the context of the EU's ongoing debt crisis. I hope to properly smack this down sometime soon.

The Grade Inflation Prisoner's Dilemma

A UNC professor that I spoke to this semester explained it thus:

"On the one hand, I don't want to give students high grades when they don't really deserve it. On the other, I don't want to punish students just for taking my class when they'd get a higher grade from another professor."

On our campus, the political science department has a reputation of being fairly tough with grades, although there is variation across professors. Sometimes students complain about low grades, and say that they would be higher in other departments. We respond that the only way to distinguish our best students is to have a fairly spread out distribution. If we give everyone higher grades, we help the mediocre students and hurt the best students. Which doesn't make sense.

As students of the prisoner's dilemma know, one way to alter the equilibrium of the game is through outside intervention. Some at UNC are trying to use administration to do just that. The thinking is that grade inflation will be less prevalent, or at least less pernicious, if grades are placed in context. A 'B' grade in a physics class with a median grade of 'C+' is more impressive than an 'A' in a humanities class with a median grade of 'A', so publishing class medians with grades changes what those grades actually mean.

As a closing aside, it seems like every semester the NY Times publishes a "state of college education" article that prominently features UNC.

Via KPC

Can Hamas Sabotage the Peace?

Economist blogger "M.S." comments on the Hamas attack in the West Bank a few days ago:

[S]omething happened yesterday that, to my recollection, has never happened before, at least not with such clarity: in the midst of direct negotiations between Israel and the Palestinian Authority, a deadly attack took place, and rather than call off the talks, both sides resolved to keep going. In fact, they both explicitly characterised the attack as an attempt to sabotage the talks, and insisted they wouldn't be sidetracked. ...

Throughout the 1980s, Israel pursued a self-destructive agenda of refusing to negotiate with Palestinians under the slogan "we will not negotiate with terrorists". Yesterday, Binyamin Netanyahu declared he would not allow terrorists to stop him from negotiating. It may not be enough to get a final status agreement. But it's an indication that he's serious.

Are M.S.'s impressions correct? Mostly. From a classic 2002 study by Andrew Kydd and Barbara Walter on Israeli-Palestinian peace negotiations (I couldn't find an ungated version):

This story illustrates two intriguing patterns in extremist violence. First, most extremist violence is not indiscriminate or irrational as many people have assumed. In fact, terrorist attacks in Northern Ireland and the Middle East over the last ten years show a clear and recurring pattern, where violence is timed to coincide with major events in a peace process. ... Second, extremists are surprisingly successful in bringing down peace processes if they so desire. Between 1988 and 1998, fourteen peace agreements were signed between combatants engaged in civil war. If terrorist violence occurred during negotiations, just one in four treaties (25 percent) were put into effect. However, if terrorist violence did not occur, six out of ten treaties (60 percent) were implemented. Although extremists are known to espouse radical views and to represent only a small minority of citizens, they are surprisingly successful in their aims.

Applying the Kydd/Walter story to the present situation would go like this: Hamas attacked Israelis on the eve of the negotiations in order to persuade moderate Israelis that moderate Palestinians would not be able to uphold any agreement made to control extremists. If Israelis cannot trust Palestinians to control extremists, then an agreement is worthless. If an agreement is worthless, then why make concessions to reach it? Hence, the attack leads to a breakdown in negotiations by causing doubt that an agreement can be enforced.

So why hasn't it worked this time? Well, it could be that neither side had any expectations that this meeting would produce a meaningful agreement in the first place. Perhaps because Israelis already doubted whether Abbas (a member of Fatah) could control Hamas extremists, so this attack does nothing to shift expectations. Both are more likely than Israelis suddenly choosing to trust Abbas in spite of the attack, and either is cause for skepticism that these meetings will lead to meaningful agreement. So while it is true that it is somewhat rare for negotiations to continue after an attack, it is not clear that M.S. is correct to view the continuance of negotiations as a positive sign. It's more likely that it's not.

The Effect of Lobbying

UNC poli-sci professor Frank Baumgartner had his research on lobbying featured in Miller-McCune (then NYTimes, then Monkey Cage):

The real outcome of most lobbying — in fact, its greatest success — is the achievement of nothing, the maintenance of the status quo. “Sixty percent of the time, nothing happens,” says Frank Baumgartner, one author of the book and a political science professor at the University of North Carolina at Chapel Hill. “What we see is gridlock and successful stalemating of proposals, with occasional breakthroughs. We see a pattern of no change, no change and no change — and then some huge reform.”

But those large reforms — such as health care for 32 million uninsured Americans under President Barack Obama, the scheduled phase-out of the estate tax under President George W. Bush, and the normalization of trade relations with China under President Bill Clinton — are far more often linked to a change in who inhabits the White House than to campaign contributions or K Street hires.

The weak link between money and policy change is counterintuitive but understandable, the authors say. The balance of power in Washington already hugely favors the rich. The status quo reflects the considerable advantages the wealthy have managed to secure in the law, down through the generations.

“If they really wanted something, they probably already have it,” Baumgartner says.

Lee Drutman has a somewhat different interpretation:

But, to me, 40 percent is actually an astonishing success rate.

Sure, this may not look like much if your starting assumption is that special interests own Washington, and that all a clever lobbyist needs to do is approach a Congressman with the promise of a campaign check and that poor helpless Congressman will practically be begging to fete that lobbyist with most indefensible corporate giveaway.

But, on the other hand, if you’ve spent any time in Washington, and you know how hard it is to get just about anything done, 40 percent is definite batting champion territory.

And the big point of the study is actually about the difficulty of change: the status quo is really, really sticky in Washington, in good part because on most important issues there are forces mobilized on both sides, and every action on one side provokes an equal but opposite reaction on the other side. Forces fight each other to stalemate for years. But then then, suddenly, there is movement – and whoever has won the war of positioning is likely to win the war of motion.

But the problem is that nobody – not even the cleverest of lobbyists – really knows which ideas and issues are likely to break and when.

The two views are not mutually-exclusive.

Politics Is (Usually) Not a Coordination Game

Matthew Yglesias sees a bipartisan rationale for overhauling the American political system:

In Canada, their health care system is more equitable and it spends less taxpayer cash per patient. Traditionally, conservatives don’t point to Canada’s health care system as a conservative victory but they do point to Canada’s low taxes which are in part a consequence of its efficient single-payer health care system. So is that a conservative win or a liberal one? Well, it’s both. The main upshot of many features of the US political system that I don’t like is to enhance the influence of interest groups and decrease the influence of ideologues and technocrats. This is basically by design and reflects 18th century state of the art thinking about the dangers of liberal governance being trampled by demagogues. Insofar as the balance we’re currently striking is inappropriate to the conditions of the 21st century United States that’s bad for the right and the left.

This assumes, of course, that politicians are primarily motivated by ideology, and therefore mutually-beneficial common ground is easy to find. Unfortunately for Yglesias, if that were true it would (most likely) have already happened, and he'd have nothing to complain about. His complaint that interest groups are able to successfully lobby policymakers indicates that his assumptions are probably wrong.

But even if the assumptions weren't wrong, his conclusion still would be. Yglesias cites the Canadian health care system as an example of positive-sum policy: both liberals and conservatives can be happy about the outcome. But is that so? Low taxes are not contingent on an efficient single-payer system. After all, taxes could be lower still without any public health care system at all, as I imagine was the case before Canada's universal health care system went into effect. Conversely, the health care system could be even more egalitarian if taxes were higher or more progressive. Any tilt in policy in one direction or the other is generally zero-sum*.

Moreover, even if efficiency gains were created by a new policy, the surplus still has to be distributed somehow. That distribution is by definition zero-sum.

The point is that it's not enough to simply point to some policy outcome and claim it is positive-sum. That judgment must be made relative to some baseline scenario.

*This can get a little bit complicated, but an exception to this rule could occur when there are scale returns at that particular policy margin. But partisans are unlikely to value those scale effects in the same way; e.g. conservatives probably don't want a more efficient public health care sector if the result is a crowding-out of private health providers. Liberals probably don't want a more egalitarian system that leaves all consumers subject to a private monopoly. Etc.

I am reminded of the urban planner Christopher Alexander and his beautiful A Pattern Language: social science is not about rules, but rather about sets of principles which can guide decisionmaking. That is, social science is about finding general patterns which help us think about specific situations. This, as I like to discuss on this site, is very different from the scientific method of the hard sciences.

Ostrom’s empirical method strikes me as the right one. Empirics do not tell us which theory to develop. Theory does not tell us which empirics to examine. Rather, theory and data develop together, feeding back on each other, in order to help us find the patterns above.

That is from a discussion of Elinor Ostrom's career, and especially her expanded Nobel lecture published in the June AER, at A Fine Theorem. There is more at the link.

RIP, David Blackwell

David Blackwell, pioneer in probability theory, game theory, and other mathematics, has died. A truly remarkable man, from a poor Midwestern town (close to where I grew up, incidentally). First black tenured faculty member at UC-Berkeley (after being turned down at Princeton), first black member of the National Academy of Sciences, and an early applier of mathematical game theory to conflict situations at the RAND Corporation. RIP.

How to Out-Crazy Kim Jong Il

Drezner notes that North Korea is playing chicken, and Carpenter asks how we can out-crazy the craziest regime on earth (they diavlog about it here):

If this is actually a game of chicken as Dan argues, how might the policy dilemma be framed in such a way that North Korea, who actually wants to avoid war, might start to believe that it's not the craziest party in the equation anymore or the one with the least to lose?

Easy. Fire Gates and appoint George W. Bush as Secretary of Defense, and pass a constitutional amendment giving him power to do whatever he wants. Game. Set. Match.

Modeling Britain's New Government

Who says formal theory has no relevance to real life? Apparently Liberal Democrats have been using game theory to model potential negotiations over the formation of a government for at least several months. After the election, but before Cameron and Clegg came to terms, Tony Price tried to model those negotiations:

Game theory says you need to understand your "outside options": what the alternative to a deal with this group is. How well you do depends on how good you can make all the outside options. Here's a go at the decision tree Clegg faces. ...

The logic of my version is:

- A deal with Labour now is unlikely to deliver PR, which is what would make it worthwhile from LD's point of view
- A deal with Tories will not deliver PR
- A Tory minority government risks giving Tories a majority in 9 months and missing the historic moment
- Clegg will be very tempted to do a deal with the Tories with inbuilt failure so that LD's can return to a PR negotiation with other parties in 9 months
- Present that deal as stesman-like: "The economic crisis demands it, and I will not make the country suffer for PR ... but once the economy is on the mend, I will insist on my PR reward"

Price came up with the game tree below.



Not bad, as it has turned out.

UPDATE: Apparently the embed of the game tree isn't working perfectly. It is here.

(ht: @TimHarford)

A Little Light Game Theory (Greek Sovereign Debt Edition)

A few days ago I saw Jeffrey Friedman extend his "Basel thesis" -- in which the risk-weighting scheme in the Basel Accords created the incentive structure that led to the subprime financial crisis -- to Greece's debt crisis:

So why did the bursting of the asset bubble in housing cause a banking crisis, freezing interbank lending and then bank lending into the "real" economy?

Because, according to the Basel thesis, Basel I bank-capital regulations, enhanced in 2001 in the United States by the Recourse Rule, encouraged banks worldwide and especially in the United States to leverage into asset-backed securities, including mortgage-backed securities, that were either government guaranteed (by Fan or Fred) or were privately issued but had an AA or AAA rating. How did the Basel rules encourage this? By giving such securities a 20 percent risk weight.

Translation: An AAA-rated mortgage backed security worth $100 required only $2 in bank capital at the 8 percent Basel rate for adequately capitalized banks. $100 x .08 x .20 (the 20 percent risk weight assigned to asset-backed securities by the Recourse Rule) = $2. By contrast, a commercial loan of $100 required $8 of bank capital, because Basel gave such loans a 100 percent risk weight. $100 x 8 percent x 1.00 = $8. Similarly, a $100 whole mortgage retained by the bank required $4 of capital, because the Basel risk weight for unsecuritized mortgages was 50 percent. With these risk weightings, securitized mortgage-backed debt offered significant capital relief.

Today's FT brings the news that "European financial institutions have $235 billion worth of claims on Greek debt, most of which is thought to be in government bonds." Why do they hold so much Greek government debt? Because the only category of bank asset treated more kindly by the Basel rules than asset-backed securities is government debt, which has a zero risk weight. I.e., no bank capital need be used to buy a government bond.

So today I was interested to read that some major banks are fanning the flames engulfing Greece:

Bets by some of the same banks that helped Greece shroud its mounting debts may actually now be pushing the nation closer to the brink of financial ruin. ...

As Greece’s financial condition has worsened, undermining the euro, the role of Goldman Sachs and other major banks in masking the true extent of the country’s problems has drawn criticism from European leaders. But even before that issue became apparent, a little-known company backed by Goldman, JP Morgan Chase and about a dozen other banks had created an index that enabled market players to bet on whether Greece and other European nations would go bust.

Last September, the company, the Markit Group of London, introduced the iTraxx SovX Western Europe index, which is based on such swaps and let traders gamble on Greece shortly before the crisis. Such derivatives have assumed an outsize role in Europe’s debt crisis, as traders focus on their daily gyrations. ...

A result, some traders say, is a vicious circle. As banks and others rush into these swaps, the cost of insuring Greece’s debt rises. Alarmed by that bearish signal, bond investors then shun Greek bonds, making it harder for the country to borrow. That, in turn, adds to the anxiety — and the whole thing starts over again.

How can we square this circle? If Friedman is right, then banks are highly leveraged in the sovereign debt of Greece (and other countries). Basel rules required a 0% right weight for any OECD sovereign debt, but not all sovereign debt paid the same yield. Some states, like Greece, are relatively more risky than others, like the U.S., but they all had the same risk weight. So banks looking for a bigger profit would plow funds into the riskier countries at a higher interest rate because yields were higher.

Why, then, would many of the same banks now do their best to increase the risk of a Greek default? If that happens, and Friedman is correct that many banks are leveraged to the hilt on Greek debt, then they lose a lot of money. They can't be trading on moral hazard, since if they believed that Greece will eventually be bailed out and their debts made whole they wouldn't waste money purchasing insurance (a.k.a. credit default swaps). So what's going on?

One explanation is that a need for hedging has created the equivalent of a bank run in CDS markets, and this is creating a self-fulfilling prophecy. Another is that this represents a classic Prisoner's Dilemma: in aggregate all banks would be better off if none of them bid up the prices of CDS on Greek debt and thus relaxed the credit constraints on Greece, but each individual bank is incentivized to defect and insure themselves against potential default. Banks in competition against each other cannot credibly commit to cooperate, so (Defect, Defect) is a dominant strategy for all banks exposed to Greek debt. This creates a run, which manifests itself in CDS markets, and leads to a sub-optimal outcome for all involved.

If this is the appropriate model, then there would seemingly be a role for outside players to influence the game. Germany, or the ECB, or the IMF, or even the US could step in and provide financing for Greece to roll over their debt, meet counterparty obligations, and loosen the constraints that Greece faces in credit markets. But this simply raises another Prisoner's Dilemma: how could a third party guarantor be sure that Greece will not defect from that agreement and continue in its fiscal profligacy? Again, the dominant strategy seems to be (Defect, Defect) unless Greece can somehow credibly commit to austerity in order to meet its obligations. Unfortunately, it doesn't appear that they can.

Maybe the EMU should play a Grim Trigger strategy: they'll provide financing for Greece in exchange for austerity. If Greece defects, then the EMU boots Greece out of the monetary union and they're on their own. That threat might be significant enough to escape the Prisoner's Dilemma if it's perceived as being credible.

Or maybe not. It's a mess. As Carlo Bastasin says over at Baseline Scenario: "You cannot imagine really solving the Greek imbalance without – at least somewhat – correcting the German imbalance." Germany does not want to correct its imbalance. So Greece is probably screwed.

(edited for correct attribution, 8:41)

UPDATE: Felix Salmon sees this as a simple hedge. He's probably right.

The Strategic Logic of American Senators

Ben Nelson gets it:

Senator Ben Nelson was the 60th vote in the Senate on healthcare reform. He held out for what is now called the “Nebraska purchase,” a side-payment from the federal government to Nebraska to extend Medicaid. He also got rid of the public option. He compromised a bit on abortion so he did not get all his cake.

There were a whole bunch of people who didn’t like the Nelson language – they only went along with because I could be the 60th vote. Leverage increases, exponentially, like the difference between a number 2 earthquake, 3 earthquake, 4 earthquake – goes up exponentially like that – your leverage goes like that at the very end…..if you are going to match Stupak, you match him at the end when you have the most leverage.

Much more at the link, including the fact that Nelson's strategy contradicts game theory.

The Prisoner's Dilemma in Banking

Felix Salmon passes this along:

“A California Banker” writes to Mish, giving yet another reason why banks aren’t lending:

If you’re a bank with a relatively healthy balance sheet with adequate capital, (like us)you want to maintain surplus capital in order to stay on the FDIC’s list of banks they can transfer the loans and deposits from a failed institution into.

This is a home run for the acquiring bank and far more of an instant benefit than any new lending.

The problem here is that healthy banks end up competing with each other to have the largest capital surplus and therefore the greatest chance of being anointed in this manner by the FDIC. If everybody was lending, the FDIC would still have to place failed banks’ assets and deposits with someone. But instead we get the opposite corner solution, where nobody is lending — except, presumably, for banks which are close to failure and need all the interest income they can get. I wonder whether the FDIC has anybody thinking about how to counteract this syndrome.

Salmon calls this the "FDIC lottery" but I think a better name would be "Vulture-Banking": the healthy banks are waiting for the sick ones to die so they can acquire their assets at fire-sale prices.

It would be easy to counteract this syndrome: start taxing bank reserves instead of paying interest on them. But the authorities seem to be more interested in the short run in capitalizing the banking sector rather than really getting cash moving again. Why? Perhaps another post from Salmon could provide an answer:

My feeling is that the US poses at least as much of a risk to the global economy as southern Europe does. There’s a good chance that 2010 could be the year of walking away from underwater mortgages; there’s no sign of the private sector releveraging; and the government has clearly reached its limit in terms of the degree it can step in and borrow on behalf of the rest of us. If the attempt to prop up the still-overvalued housing market fails and there’s another downwards lurch, there will be a whole new wave of bank insolvencies and much less fiscal space to bail them out than there was pre-crisis.

Right. So if the regulatory authorities are thinking the same thing, they want to make sure there is enough capital in the banking system to keep banks solvent if there is another wave of writedowns in real estate. And if/when more banks do collapse, they want to make sure that other banks are healthy enough to absorb their balance sheet. The government either can't or won't pass another TARP-type bill, so the strength of the banking sector is essential; there can't be another bailout.

The downside to this strategy occurs when a cessation of lending slows down the economy enough to cause another downturn, which causes more foreclosures and walk-aways, which pushes more banks into insolvency. The regulatory authorities, however, apparently think that is a risk worth taking.

UPDATE: McMegan says this is evidence of moral hazard. I don't think so. I think regulators would like to see banks lending more, but they also want to boost capital reserves as protection. But they've incentivized the latter, not the former, and banks have responded accordingly.

Gaming Chinese Marriages

Tyler Cowen passes along this little nugget from China, where a husband and wife entered into a contract whereby she may only beat him one each week.:

The 32-year-old man, who was named by the Chongqing Evening News as Mr Zhang, took the unusual step after suffering intense abuse from his wife, who studies kung fu.

"I don't want to beat him, but arguments are inevitable and I can't help myself," his wife told the newspaper. She added that in the week before they signed the deal, she had beaten him up three times. ...

In order to curb his wife's aggression, Mr Zhang proposed signing a contract in front of his in-laws. If his wife breaches the contract, she has to return to her parents' home for three days. "She is very obedient to her parents, and her parents will support me and blame her," he said.

I was a bit surprised that Cowen didn't tie this to the already-bad-and-worsening highly skewed female-to-male ratio in China, which is largely a product of the one-child policy but has the side-effect of creating a scarcity of women. This provides the existing women with lots of relationship leverage, which might be why Mr. Zhang's parents are relatively nonplussed:

Mr Zhang's parents told the newspaper that although they felt bad that their son was regularly attacked, the couple were a good match. "They have a good marriage, so we can say nothing about it," said his father.

My question is this: how will this really affect the marriage dynamics in the Zhang household? Cowen also neglected to discuss this, which is very un-Tyler-like. I can think of a few things:

1. If Mrs. Zhang uses up her beating early in the week, then Mr. Zhang can act with impunity until the start of the new week. Therefore, Mrs. Zhang should use her beating as leverage and save it until later.

2. Knowing this, Mr. Zhang should push her buttons early in the week. If she succumbs and beats him, then he can behave however he likes for the remaining days. If she reserves her beating, then he gets his way.

3. The closer it gets to the end of the week, the less of a deterrent the threat of the beating becomes. Why? Because Mr. Zhang knows that it's coming eventually, so he may as well act as he prefers. By the time the beating already comes, Mr. Zhang will have had a full week's worth of antagonism. Presumably, he values this or they wouldn't be in this situation in the first place.

In other words, despite being stronger, Mrs. Zhang has lost all leverage in the relationship and Mr. Zhang is now incentivized to behave badly. This assumes, of course, that Mrs. Zhang strongly prefers not to live with her parents. If she does not, then she may as well beat Mr. Zhang twice on Monday, come back on Thursday and beat him again, and spend the weekend at her folks' place. Lather, rinse, repeat.

Or, if you prefer: wax on, wax off.

But even if she prefers to live with her husband rather than her parents, she should still beat him twice on Monday if he misbehaves at all to prevent his misbehavior the rest of the week, so long as she prefers her parents' house to his shenanigans. This can change the equilibrium if Mr. Zhang prefers living with his wife even if it means compromise over being beaten repeatedly and living alone. So he may choose not to misbehave at all. In this case, the deterrent isn't the beating, but the resulting separation. All the leverage shifts back to Mrs. Zhang: by tying her hands, she is getting what she wants without having to beat her husband. Schelling would be proud.

Of course if that was the marriage dynamic, Mrs. Zhang could've just threatened to go to her parents' house whenever Mr. Zhang upset her. But then they would never have been in this situation in the first place. So I'm sticking with my first interpretation, and predicting that the marriage does not last.

Tiger Woods, call your lawyer. This is better than the deal you're getting right now.