Saturday, November 19, 2011
Saturday, November 12, 2011
Monday, May 02, 2011
Economics against corruptions – some random thoughts
The (cricket) world cup over, a feeble rally in SENSEX, a run-down IPL and everything else is business as usual …boring!! Of course, that was before the recent waves of anti-corruption movements, that is. Was it some event, a septuagenarian fasting unto death, hundreds of people thronging the streets, thousands of pages in support on social network sites. All this to pass a law – against corruptions – that is apparently running late by roughly 42 years! Last similar event in nature and scale I remember was the 2006 anti-reservation thingy. It went nowhere.
One of the things about corruption is people do not seem to like it. The other thing - it is mostly a rational behaviour. Economically speaking, it is mostly just rent-seeking. A government having the monopoly over natural resources sells a permit at a price of P. The government official, in charge of selling the permit and having monopoly over his own signature, puts a mark-up and sells it at P1, the difference is the bribe. For an end consumer, the effect is just a price increase. Ethics and morality aside, does it really make a difference? Should you pay the bribe and/ or try stop corruptions? The answer depends
Paying the bribe is almost always rational if the gain you make by doing so, less the bribe amount, is more than what you have otherwise. Be it a personal transaction like applying for a passport, or a business transaction, like applying for a forest clearance. But whether you want to stop it? The answer is quite different in these two cases. For a passport, if you do not have to pay the bribe, that increases your net utility, so you would ideally like it to stop. But for forest clearance, an absence of bribe might also mean more competition lured in to your business, and hence a lower profit. So not necessarily a company paying bribe would like it to stop in an oligopoly structure. Unfortunately, when we have more chances that agents are against bribes (the passport case), the incentives (the amount you pay, or save otherwise, compared to the effort to stop it) are pretty low, and when money involved is really high, there might be no much incentive to stop it. That is why without explicit policy instrument, corruption can be hard to kill. We will have to wait and see how inspired the people are beyond economic rationality to make the structure of the Lokpal Bill a success
But should even policies be aimed at stopping corruptions? In an abstract sense, bribe is just a transfer of wealth, somewhat similar to tax. If there are complicated rules and procedures, paying bribes can help avoid red-tapes and increase efficiencies, and can actually be good for the public. But on the other hand, bribes might also incentivize more complicated rules creations, thus hampering efficiency. In general people tend to believe the net effect is indeed negative. Plus on an aggregate level, bribes also has some dead-weight loss (like for the passport case above, the net increase in utility for the bribe-takers will be less than the loss of that of the bribe payers, total utility goes down, which economists think a bad thing to have). And also it is somewhat discomforting that bribes also means corrupt people gain at the expense of honest people (we have many income redistributions in modern society, like reservations mentioned above, but this is perhaps quite hard to justify).
So once we generally agree yes, this is a bad thing, then comes the next question, how to stop/ reduce corruption. Let’s, just for fun, check out some simple stuff. Assume in a corrupt transaction, the official demands a bribe b, and if caught, with a probability of p, will incur a cost of m1 (loss of job, reputation, punishments etc). Also assume the bribe payer will have an excess gain g, if he pays the bribe, and also, if caught, incurs a loss of m2. So the utility functions of the bribe taker is x1 = (1-p)b – pm1, and for the bribe payer, it is x2 = (1-p)(g-b) – pm2. Now they bargain, and if they are smart, they settle down at b = 0.5[g + p(m1 – m2)/(1-p)]. This implies that bribing happens only when m1+m2 < (1-p)g/p. This leads to some expected results. Like, chance of bribing goes down with increasing punishments (m1 and m2), increasing probability of detection (p) and decreasing gain (g). But perhaps there are more hidden insights in this simple model. For the case of passport, g is low, and since bribe payer has incentives to stop corruptions, p will increase as m2 is reduced (like in extreme case, a reward for the bribe payer for reporting a bribery case). Under this scenarios, reducing g (i.e. making processes more efficient to reduce the pain of going the straight way and not paying the bribe) and m2 (reward for reporting a bribe) are more effective, than increasing m1 (punishments) or p exogenously (high cost of monitoring of a large number of small transactions). But in the scenario of the forest clearance, things are different. There m2 does not affect p (as briber has incentives to continue the corruptions) and p is probably dependent on g (a large scam is more likely to be detected). There raising the punishment (m1 and m2) is most effective, along with close monitoring (exogenously increasing p). So in this context, the proposed bill only addresses to m1 (and perhaps m2) and p (exogenously, by increasing monitoring), and at best will be a partial success. It is important to have close monitoring and grievance redressal system, but perhaps more important to take notice of the underlying incentives
Wednesday, December 29, 2010
Twenty-Ten, Close of Business
So here are the things and events that worried the world this year. Surprisingly, people didn’t care much about the major events in the finance and economy – like the sovereign crisis or the Wall Street reforms or the currency wars. Seems like the banker bashing sentiments is a myth created by the media?

Back in

And out there in the deep space, yet earthlings remained pathetically limited. Still wandering around the moon or the Venus or at most Saturn. The best achievement we could manage was some freaky scientists testing if the universe is really a 3D holographic illusion (like Matrix? WTF). And all this while nobody seemed to care about the sudden disappearance of some of the characteristic stripes on the Jupiter

And finally, coming back to

____________________________________________________________________
disclaimer: graphs are based on data mostly from Google Trends, and other web resources whichever I found convenient
Sunday, June 20, 2010
World Cup 2010 - Number Crunching
Here is a quick and wasteful analysis of the world cup outcome possibilities. Given the results so far (till Cameroon Denmark last night), here are possible lists of group of 16 qualifiers, and in-group rankings based on 3 scenarios to decide the outcome of the rest of the matches in 1st round : 1. Fifa Ranking (better ranking team wins) 2. Momentum (the current group order stays) 3. Most likely – based on my opinions + betting odds implied probabilities
Group A: Almost surely ->
Group B: Almost surely ->
Group C: The fight appears to be among
Group D: Again high uncertainty,
Group E:
Group F: Almost surely
Group G: Almost surely
Group H: Fight between
So from
The betting markets in the meantime are showing some odd probabilities, mostly influenced by English and EPL fans IMHO J . It says, given they qualify for it, Spain has a chance of surviving (22%) the knockout rounds better than Argentina (20%) or Brazil (19%). Same for
And finally here are the probabilities of the winning country, again based on betting odds –
Note: All odds are taken from Betfred, other data from FIFA. For those interested in running their own simulations – here is a nice excel (http://rapidshare.com/files/398535274/World_Cup_2010.xls). Please note, this is not made by me, I used it, and did not find anything problematic/ harmful, but I cannot promise so
Tuesday, January 12, 2010
Tonight I Can Write the Saddest Lines (HHGTTG Version)
Tonight I can write the saddest lines.
Write, for example,'The earth is shattered
and the great yellow spaceships hang motionless in the distance.'
The photon storm revolves in the sky and sings.
Tonight I can write the saddest lines.
I loved her, and in some spacetimes she loved me too.
Through nights like this one I held her wrapped in my towel
I kissed her again and again under the endless sky.
She loved me in some spacetimes, and I loved her too.
How could one not have loved her great still eyes.
Tonight I can write the saddest lines.
To think that I do not have her. To feel that I have lost her in temporal realsitics.
To hear the immense night, still more immense without her.
And the verse falls to the soul like Pan Galactic Gargle Blaster to the brain.
What does it matter that my Information Illusion could not keep her.
The Time has ended and she is not with me.
This is all. At the End of the Universe someone is singing. In the Restaurant.
My diodes are not satisfied that they have lost her.
My disembodied mind searches for her as though to go to her.
My Heart of Gold looks for her, and she is not with me.
The same old folks spending another year dead – for same reason.
We, of that spacetime, are no longer the same.
I no longer love her, that's not entirely unlikely, but how I loved her.
My voice tried to find the wind to touch her hearing.
Another's. She will be Somebody Else’s Problem. Like my kisses before.
Her voice. Her bright body. Her infinite eyes
I no longer love her, that's not entirely unlikely, but maybe I love her.
Love is mostly harmless, forgetting is another thing
Because through nights like this one I held her wrapped in my towel
my diodes are not satisfied that they have lost her.
Though this be the last panic that she makes me suffer
and these the last verses that I type for her.
If you liked it, great. If you didn’t, DON’T PANIC – as I mentioned already above, I will try my best not to write any more verses – “for her” or for any goddam body, thing, or particle for that matter.
Saturday, July 11, 2009
Inflations? What Inflation?


