Though the Parties may have moved in a rather unexpected path to reach an understanding with the Masoists, the so-called alliance has an inherent risk of collapsing sooner than later.

The parties may need to do something more dramatic, may be just to bluff the Palace if for nothing else.

A very interesting article by Ashutosh Tiwari.

source: - http://www.nepalitimes.com/issue276/strictly_business.htm

The article may not follow the orthodox political path that many Sajhaites may be accustomed to. But thanks to the author's wide spectrum of knowledge and commonsense, this artilce is very interesting and equally persuasive.

Nice piece and makes sense too.
However, is it possible that the act is already in progress, like maoists realizing the need to end the violence that itself is in danger of going out of control. The parties and palace sensing that the best option is to acomodate each other in the political context of Nepal and now having to give due space to the third front. The bipolar power struggle in the current scenario is likely possible to remain so, with current alies spliting on ideological fronts to be the contenders of power, with the current one power center opting to play neutral and a balancer.

The political parties in Nepal are making a big mistake.The only way to resolve the problem in Nepal is by exterminating even the last of the maos.The alliance does have its short term benifits but in a long run it will be a move to make maos the sole political power of Nepal.As of now the political parties will be successful in pressurizing the king to restore multiparty democratic system.But, in the long run this alliance will be only beneficial to Maos.Nepal needs to get rid of both maos and the King.There can't be a peaceful resolution to this problem.I hate to say it but the bitter truth is a big bloodbath is necessary to get rid of both the king and the maos.

hey bhoonte,

I was looking for u everywhere, here u r.

Have u changed track from being royalist now. I am just wondering. I was rofl reading ur list of "real" nepali heros.

And btw yes the article on nepalitime.com was interesting and "different".

Credible indeed.

Hmm, so if one has to act crazy sometimes to be credible, it is only natural that the other party expects her to act crazy sometimes just for the sake of being credible. In that case, does acting crazy contribute to credibility?

In other words, as the old adage goes, "Expect the Unexpected." But, if you expect the unexpected, is the unexpected still unexpected?

blah...

Trying to be balancing and realistic to the situation but it gives me the impression that the author is being sympathetic to the current regime. But then again thats his right. All are free to think the way they want, which is the sweetness of democracy & freedom... :p

Biswa, I was always a pragmatist. I am always open to ideas. If you think that that makes me a communist, or royalist or rightist or whatever, I do not have anything to say to your petty comments.

I am always for peace and prosperity of Nepal and thus adovcate for taming of terrotists and taming of UML and Nepali Congress before handing any executive power to them.

Here's an agreeable criticism.

Though one can't cover as much in a 600-700-word-piece, which has to make one clear point (to stimulate readers' thought) and leave it at that, getting critical comments -- even when is NEITHER a writer NOR a journalist -- nonetheless opens up ways for one to learn, reevaluate and think deeper about one's arguments in more nuanced ways.

That's the beauty of writing openly and publicly.

Obviously, I need to sharpen my arguments in weeks ahead to take account of these (and other) criticisms (that friends have made in private emails.)

Enjoy,

oohi
ashu

**************

Crazy

Although Ashutosh Tiwari’s attempt in his Strictly Business column to apply game theory to the political gridlock between the King, the Maoists and the political parties (‘Crazy as a strategy’, #276) is an admirable step in beginning a more rigorous discussion of the political crisis in Nepal, his analysis is inadequate on two counts.

First, he recommends that the parties play irrational strategies to make future threats credible but he fails to recognise that the political parties together do not form a single collective institution that is opaque in the way the Maoists and palace are. Their agenda is developed in a discursive manner (as the open publication of Tiwari’s own article suggests).

So their strategy to play some irrational strategies for the sake of making other threats credible would be common knowledge both to the Maoists and the monarch, whose responses would take that knowledge into account, resulting in suboptimal outcomes for the political parties. Therefore, without a more robust mechanism for making threats it is unlikely that Tiwari’s recommendations will be useful. (They may even be counter-productive.)

Second, Tiwari accuses the parties of ‘never pose[ing] a credible threat to the palace and the Maoists’ but it is not clear that they were ever in a position to do so. He suggests replacing ‘old politicians with stridently republican ones’, but does not offer a mechanism by which to do this.

Since intra-party games determine which individuals emerge as the leaders, what is required first is a more in-depth analysis of leadership evolution within parties. Further, it is unlikely that ‘call[ing] the palace’s bluff’ and ‘tak[ing] part in internationally supervised elections’, is even a feasible strategy. I am not convinced that foreign countries or organizations would be willing and able to supervise elections in Nepal (and recognise their outcomes) without the palace’s consent.

Tiwari is right, however, that for the parties to actually succeed they need to be more creative. The application of a well-known result in bargaining theory–that the bargaining outcome depends on the disagreement payoffs (ie: what each party gets if negotiations break down)–tells us that because the parties do not receive much in a stalemate or breakdown of talks, they do not have many bargaining chips to work with.

Given these shortcomings, it is my conjecture that a more sound game theoretic analysis that seeks to maximise party payoffs would construe well-defined mechanisms for cooperation between the parties and the Maoists, or unity between the parties and the palace. I am inclined to believe that unity between the parties and the palace would yield better payoffs for the country and for the palace, as opposed to the development of the parties-Maoist alliance that we are witnessing now.

Avidit Acharya,
email

Source:
- http://www.nepalitimes.com/issue277/letters.htm

It does not take Schelling to see that bringing staunch republicans in the leadership will make our political parties stronger.

The strength of political parties is the mass and recent demonstrations have clearly shown that the mass rallies only behind the slogan of republic. And let us not forget that the Maoists came to this point of their strength riding the wagon of republic.

However, Ashu's conclusion that political parties can strengthen themselves by taking part in the election announced by the King is ridiculous. I doubt one can reach to that conclusion by applying Schelling's theory.

I do not know much about Schelling's theory. So, only Ashu can tell whether he actually applied Schelling's theory to reach to the conclusion or he simply mentioned the theory to sound credible while serving his own incredible theory.

Or somebody who is familiar with Schelling's theory.

Here is one such person [name withheld rey kya] . This gentleman is a student of the Game Theory and his conclusion is quite different than Ashu's. I am taking liberty to post the theoretical part of the gentleman's argument shared in Nepal Democracy Google Group.

Here it goes:

....My result, not surprisingly, is different from [Ashu's]. It uses elementary statistics and ordinary differential equation of first order.

Model:

The politics of Nepal after the parties Maoist pact can be defined as a two player game, in which there are two payoffs:{election under king for municipalities, election for constituent assembly}. Let it be considered as a standard war of attrition game in which two players make offer (2,1) alternately. If one refuses, then it is another person’s turn to make offer of same (2,1). Here, we can interpret 2 as constitutional assembly election for parties/Maoist pact, and 1 as election under king. For king, this is the other way.

After the pact, the king understand that there is a probability p that the coalition of his opponent be coopted by the Maoist agenda, i.e. never gives up in its demand for constitutional assembly.

Both player understand that the king is likely to give up at some finite time T (No one can rule forever, afterall). If it is sufficiently late, then former coalition might think it will be better off taking part in the election. Let F1(t) be the cumulative density function for the king’s surrender, such that there is a time T which is finite and at which F1(T)=1. As for coalition, if it goes Maoist’s way, it will never concede, but if it is prevailed by the parties, then there is a cumulative density function F2(t) of their giving up. We also assume that their discout rate is same, r.

Result: In Bayesian Equilibrium, the alliance will not give up before time T (i.e. the time at which the king concedes with probability 1). Furthermore, in Bayesian Nash Equilibrium, it is strictly better off for the king to give up immediately: i.e. the king is better off not trying to hold on to the power.

Proof: The proof has some ministeps. First, it should be obvious that the parties will not give up with probability 1 before time T. Why? Because suppose we think they will give up before T, at some point we call T’. Now, if they don’t give up at T’, the king will think that the alliance has coopted the Maoists agenda and will surrender immediately,which contradicts our assumption that T’
Also, note that there are no subintervals in [0,T] at which both the parties and the king has flat cumulative distribution functions, nor can they put a positive weight at any point, creating some jump in their cdf. In fact, both the distribution functions must be strictly increasing and continuous. (This argument requires a bit more formality, but is easy to prove.)

Finally, some ugly math. With a continous and strictly increasing give up cumulative distribution function, the parties must have a constant payoff in the interval [0,T].

Nepe,

First, let me rant against your attitude, before I get to the mathematically sexy model.

Only a Marxist like you think of this sort of PUBLICLY held kura-kani in terms of
"who's right and who's wrong; and who's winning and who's losing in this debate".

Only the-world-is-in-either-black-or-white mindset like yours tend to cast everything
in starkly adversarial terms in a cheap effort to advance your political agenda.

[This sort of thing is, by now, EXPECTED of you. And I might as well say that I have
long STOPPED thinking of you as an intellectually honest person --- just someone
who dutifully studied science up to a point and is now neither creative nor driven to advance knowledge in an open-minded and intellectually honest manner. But that's the way you are, right?]

But those who keep an open-mind about this sort of thing WELCOME and even HUNGER FOR brutal criticisms/improvements/additions of his/her ideas so that further/deeper thinking might result.

After all, what's the point of SHARING one's thougts in public if those ideas do not make others think hard enough to invite DIFFERENCES of opinion/disagreements so that one can revisit/remodify/deepen one's own thinking?

Isn't this how knowledge advances, even in that part of New Jersey, Mr. So-called Scientist?

That said, thanks for bringing this to Sajha. The name of the person here is NOT important; his thoughts are.

After all, it's not everyday that one sees integral signs on Sajha. So chalk that up another reason to give money to Sajha. :-)

*****

That said, the model -- like all game-theoretic models -- makes sense ONLY within its NARROW, tight logic.

In order to advance kurakani in an open-minded way, let me offer some thoughts.

QUOTE: "Both player understand that the king is likely to give up at some finite time T (No one can rule forever, afterall)."

This is debatable on two counts. Yes, the present king can't rule for ever. But, in theory, he can be succeeded by his son, thus making the transition from him to the son a continuous affair, and thus elongating the time T. Alternatively, possibilities do not exist in one direction only : there exist (theoretical) possibilities that the Maoists might give up or that the parties might also give up being together with the Maoists.

QUOTE: "If it is sufficiently late, then former coalition might think it will be better off taking part in the election."

This appears to be the king's strategy -- to make everything sufficiently late.
I would assign a greater weight to this condition.

QUOTE: "Let F1(t) be the cumulative density function for the king’s surrender, such that there is a time T which is finite and at which F1(T)=1. As for coalition, if it goes Maoist’s way, it will never concede, but if it is prevailed by the parties, then there is a cumulative density function F2(t) of their giving up. We also assume that their discout rate is same, r."

Again, this is a one-track logic, which holds the party-Maoist allaince constant or casts it in a positive light, while letting everything ONLY (around the king) change. Why not leave room for the Maoists' surrender or to the collapse of the alliance -- for there exist theoretical possibilities for such options too.

To keep the math tidy, the model makes more debatable assumptions, such as: QUOTE "Now, if they don’t give up at T’, the king will think that the alliance has
coopted the Maoists agenda and will surrender immediately."

Another HUGE assumption: "Also, note that there are no subintervals in [0,T] at which both the parties and the king has flat cumulative distribution functions, nor can they put a positive weight at any point, creating some jump in their cdf."

These assumptions fail to take account of, at the very least, what my critic AA (see above) wrote: That the assumptions take the 7 parties as a monolithic entity, and thus fail to take into account of inter and intra-party dynamics that affect the strength/durability of the alliance. The alliance is not something fixed; it too is buffeted by seen and unseen forces

In summary: A good model; makes one to think. But, like most-game theory models, makes a set of assumptions and then confines the use of logic and math to stay within the boundaries created by those assumptions. Then again, game theory is an analytical tool with its limitations, and NOT a fool-proof crystal ball (as Nepe appears to make it out to be).

The model's conclusion is PRIMARILY the same as mine: "Parties must be credibly theatening to the King".

Keeping an open mind, in my NT article, I refrained from dictating the terms of that threat to the parties; while the model-maker, behind all the fancy math, lets his biases show by refusing to consider alternative scenarios and hammering only one point home: That the King must surrender. No wonder Nepe loves and hugs him.

Thanks, and feel free to disagree.

oohi
ashu

Are you guys are talking same Game Theory as in Movie Beautiful mind?

Looks like we Nepali one beautiful minds of " Sajha Buddijeevi" supporting/defending King G's move.

Damn.!!! Kano goru lai aausi na purne... this complicated mathematics!!!

cool discussion.

I will leave Ashu's cheap and false rant against me for some other time.

Here, I am bringing a response to Ashu's comment to the anonymous student's version of the application of the Game Theory to the current conflict.

Nepe
________________________________
Response from the anonymous student:

{QUOTE: "Both player understand that the king is likely to give up at some finite time T (No one can rule forever, afterall)."

This is debatable on two counts. Yes, the present king can't rule for ever. But, in theory, he can be succeeded by his son, thus making the transition from him to the son a continuous affair, and thus elongating the time T. Alternatively, possibilities do not exist in one direction only : there exist (theoretical) possibilities that the Maoists might give up or that the parties might also give up being together with the Maoists.}

I actually think the assumption gives even better result when we think about monarchy, in stead of the king. While king can’t rule forever, given statistiscal data available, one can say no dynasty can rule forever. Once we assume the king knows this, we can fairly put a date T. Let’s say T=1000 years(i.e. based on history, the king knows his dynasty can’t survive more than 1000 years). The result holds in this case with a quite good implication, that if he is looking for welfare of his dynasty, he better give up immediately which also conforms to our intuition (as long as the parties put positive probability that they will coopt the Maoist strategy).

I now discuss briefly with the theoretical possibility that the king might not give up with some positive probability, so that we can see how the results look. This is a mirror situation for a crack in the coalition, given what we have assumed so far. Fortunately, it actually explains how the politics is being run in our country rightnow in my view.

In this model, the king is quite obstinate. Suppose with probability q1 the king thinks he (or as explained above, the dynasty of Shah) will never give up, and with probability q2 the PCM(Parties Coalition With The Maoists) will never give up. Obviously, the resoluteness and intransigence of the players do matter in this case. Who should look more resolute in this case? PCM or the king? In this case, let G1 be the cdf of king, and G2 be the cdf of the PCM; when they are surrendering type. Of course, if they tend to slug it out until infinity, the whole cdf concept doesn’t work, since that infinite fighting with each other in itself is a unique Nash Equilibrium.

I try to be brief on this proof too. I urge the readers to look at the literature on the famous War Of Attrition games where these arguments are considered in elementary levels, so I consider them as obvious for brevity.

Arguments without detailed proofs: (please look any standard text books for this if they are not fairly intuitive to you. One good source is this latest paper Yossi Feinberg and Andrzej Skrzypacz (2005) “Uncertainty about Uncertainty and Delay in Bargaining.” Econometrica 73 (1) pp. 69-91.)[again, these are not assumptions]

[The discount rate gives the degree of patience of people involved.]

Now, this is the time to use the boundary conditions to find some meaningful interpretations.

To understand social science models, we have to simplify some of the assumptions first, and see how things behave in this ‘simpler’ and ‘idealized’ models. Then we can project these results into more chaotic world. This writer has neither such intellectual capability nor time to pursue it in that dimension, so I leave it to the readers to think about that case. It is scientific tradition, including in Physics and other disciplines, to consider any phenomenon with that approach, for example, earlier analysis in Physics always assumed frictionless environment to study velocities and speed in an environment. So, in a somewhat simplified world, let’s assume both the king and the PCM have same discount rate, i.e. r1=r2. I have also advised in our previous discussion that parties be more resolute, so q2>q1. This also mean that they are the one for whom G2(0)=0.

{Another HUGE assumption: "Also, note that there are no subintervals in [0,T] at which both the parties and the king has flat cumulative distribution functions, nor can they put a positive weight at any point, creating some jump in their cdf." }

Fortunately, this is not a mere assumption. This is a proven result, and often taken as a granted while doing these analysis. Please take a look at the above mentioned paper for some more intuition on this or I will write a brief proof for this should the need arises.

{The model's conclusion is PRIMARILY the same as mine: "Parties must be credibly theatening to the King".}

I only partially agree with this: the result here clearly calls for parties to NOT take part in the election, while Mr Tiwari’s one of the recommendations was to take part in the election.

Other than that, I admit the weaknesses of the model. Like all social science models, it doesn’t capture every chaos, every unpredictability and every mood of each player. But this is how all models work in economics; otherwise, we would have predicted stock market and other unpredictables long ago. The reality is, of course, we don’t know what is good for a person in an environment of uncertainty, the best we can do is predict based on a rigorous model and time honored logic. We take what tool we have as a student of science, and try to see how things are (or should be) behaving. That’s all we do. We are not jyotishi, nor are we the 'we-know-all' sages.

I hope this clarifies my position, and helps interested people see things better in a rigorous way. It is not fair for us and our readers to reach to a conclusion just because we like those conclusions.

__________________

Cool discussion... ... ...

very funny:politics->math->politics, as if all this is a joke!

self proclaimed mathematician
pire

I will be waiting to see how rigorously Ashu applied the game theory to come to his conclusion about the election. For now, a little bit old cartoon from Batshayan on the same theme.

Nepe,

I am DELIGHTED that my piece in The Nepali Times (and NOT, alas, in the Quarterly Journal of Economics) has made others think about these issues. That's really great.

My attitude is: What's the point of writing columns for general readers -- columns that aim to talk about issues at the intersection of economics, public policy and law and tries to see how they relate to contemporary Nepal -- if doing so fails to make readers think, agree or disagree with my views?

Other than that, temperamentally, I enjoy taking risks with ideas -- knowing fully
well that half the time, even if one is wrong, that will be perfectly all right, for one
can quickly process the lessons in a non-personal way and move on. The key is not
to falter but to keep on at it and take objections/criticisms into account and come up
with better models/arguments.

That, so far as I am concerned, is how NEW KNOWLEDGE continues to get generated.

In this, I draw my inspiration from the second paragraph of the following link :

-http://www.freakonomics.com/blog/2005/04/06/an-ode-to-gary-becker/

***********

But it's plain that this sort of attitude is alien to you. I mean, what do you want me to do now? Throw the latest model of Econometrica, and chest-thump by saying, "My math
is sexier than your math", and reduce this whole thing into a show of ego, in to some contest? And what cause will that that serve?

To his credit, the anonymous student talks about the role of uncertainty in game
theory. I agree with the role of UNCERTAINTY here. That means, in practical terms,
we will have to wait and see how things turn out in Nepal especially given the fluidity
of the situtaion and inter and intra-party dynamics in days ahead.

That said, I am NOT done yet. This is just a beginning. Since I am NOT going away anywhere, I will continue to write for The Nepali Times, and will continue to study
and think about economics, and I will continue to apply game-theoretic thinking to understand Nepali matters better in days ahead -- inviting criticisms, objections
and disagreements so that much learning can result.

Stay tuned.

Other than that, I hope that my pieces in The Nepali Times will continue to make others think about these and other issues. That's the whole idea.


oohi
ashu