Friday, May 20, 2011

Algebra of Democracy

It was fun time in India last week, with election results coming out for five state Assemblies. Being a political junkie, I didn’t have enough of those political commentaries, insightful interviews, breaking news and plain hot air. Now that the last political observer has gone home to roost, let me talk about a numerical election conundrum.

Let me state the problem first. Let us say there is a hypothetical tiny Indian state called Jungal Pradesh (JP) with a ten member Legislative Assembly. The state therefore has ten ‘constituencies’, each electing an MLA. Let us say there are 10 Lakh voters in this state and that the population density is so uniform that each constituency has exactly 1 Lakh people. (Ten lakh people divided into ten constituencies)

Let us also say that there are four political parties in the state – Party A with a popular support of 40% of the state population (4 Lakh supporters in all), Party B with 25% of popular vote (2.5 Lakh supporters), Party C with 20% of popular vote (2 Lakh people) and Party D garnering the remaining 15% vote (1.5 Lakh people supporting it). And I am the leader of Party B, Mr. Haraami Lal. The popular vote percentages are gleaned by opinion polls and are generally agreed to be reasonably good estimates. Question – How many MLAs will make it from each of the parties? In particular, how well will my party (Party B) fare in the elections?

First of, this is not a trivial question. As we will see, democracy works in strange ways.

Before we do data crunching let me point out that we don’t even have to have elections if we know the popular vote percentages. We can simply have ‘proportional representation’, meaning you get to send MLAs in proportion to your popular vote percentage. Party A, for example, has 40% of the popular vote and therefore it seems only fair that they get 40% share of the MLAs, which works out to be 4 MLAs. Similarly my party will get 2.5 seats, Party C 2 seats and Party D 1.5 MLAs. (since half MLAs is ridiculous, let us say that my party gets 3 seats and Party D, 1 seat. Remember, I am a slimy, sneaky politician) This arrangement seems very fair to everyone. Not only that, but since every party gets representation, we can be reasonably sure that all aspects of issues will be thrashed out and the resulting policies will be fair to everyone in the population. But the problem with this arrangement is that no party has a working majority and if you don’t build consensus, the government activities may be completely paralyzed or we may have compromise hotch-potch policies and laws. Of course, parties can form alliances and establish a majority. But the reality is that these political parties have different identities and any amount of alliance building can come apart at the seams any time.

Let us look at what happens when we have standard elections. Let us look at Scenario 1 where we assume a totally homogeneous demographics. That is, not only do 4 Lakh people vote for Party A in the entire state, each constituency in the state also votes in a similar manner. That is, each constituency contains 40000 people (40% of 1 Lakh) supporting Party A, 25000 supporting Party B, 20000 supporting Party C and 15000 supporting Party D. Now, what happens when you have a good old fashioned election? (Assume for the moment there are no prior electoral alliances between parties) Party A will win all ten seats and will land 10 MLAs. And there will be nothing for any other party. Newspapers will be full of stories about the landslide victory of Party A. There will also be no problem regarding majority. The most popular party gets to rule, which is fair. But the other parties, with a combined numerical majority of 60% get zilch in the deal, which seems so unfair. And democracy is supposed to be representational.

Let us look at scenario 2. Remember I am Mr. Haraami Lal, the leader of Party B and I am a sly politician. I will do a dirty, hypothetical trick. I will try to manipulate the demographics. I will artificially force my people to relocate in the state. I will move my 2.5 Lakh voters such that 51000 of them are now in Constituency 1, 51000 in Constituency 2, 51000 in Constituency 3, 51000 in Constituency 4 and the remaining 46000 wherever. Once I have this redistribution, if we have elections, my party, Party B, will win the first four constituencies where I have numerical majority. In other words, Party B will have 4 MLAs. Party A, with a high numerical lead can still win most of the other constituencies and get the majority and form the government. But compared to Scenario 1, where I got no seats, now I get a whopping 4 seats!! Enough to become a force to reckon with. Or at least enough to be a nuisance.

And here is a twist. The same ‘redistribution of the population’ can be done by Parties C and D as well. Party C can plant 51000 of its people in Constituencies 5, 6 and 7 and win 3 seats and Party D, similarly can distribute its sympathizers and get 2 MLAs and the poor Party A will be left with just one seat!!!

So, we see that demographics is very important. Even if you don’t have state-wide popular support, if you have local popular support there is a good chance that you will pick up seats. Inverting this argument, if you want to start a new party and win MLAs, just grab on to an issue or a problem that is worrying a majority of a local population and project yourself as the leader.

Caste is the simplest example of such an issue. If caste sentiments are rampant in your constituency, become the leader of that caste. The only problem with this is that we have so many castes in India and you cannot be the leader of all castes. And also in many parts of India, caste is becoming a non-factor. Religion, regionalism, farmer’s crisis or any of the other single-issue issues can be used to skew the demographics in your favor and win yourself elections.

Now let us look at scenario 3. Remember I am Haraami Lal, leader of Party B. I see that I only have 25% of the popular vote. But if I joined hands with Parties C and D and form an alliance, I can command a total of 60% and beat the vote leader, Party A. So, I will walk up to the leader of Party C, and suggest that we form a coalition. I will allocate him 2 seats after much horse-trading. (which works out to be 20% of the total number of seats, representative of Party C’s popular vote percentage of 20%) I will also walk up to the leader of Party D and offer him 1 seat (because this is almost equal to his popular vote of 15%.and also because I am slimy. Of course, I promise to make him a minister. If I were nice, I would have given him two seats by using a different round-off error) And I will contest in seven constituencies. Because the coalition’s total percentage of popular vote is 60% we will sweep the elections (assuming homogeneous demographics) Newspapers will rave about our ‘clean sweep’ and I will win seven seats and get the majority strength. And remember, I am the slimy politician Haraami Lal. So, two days after the elections I will dump Party C and Party D and rule by myself. From zero seats under scenario 1 to seven seats under scenario 3. I have leveraged off my 25% vote into a majority!!!

Scary, isn’t it? By just doing a few tricks – dirty ones, no doubt - and by forming suitable alliances you can really run for the roses.

The obvious conclusion is that pre-election coalition politics – the multi-partner marriage of convenience is absolutely harmful to democracy. We got to have very strict laws in place to check its game changing ways. It serves absolutely no democratic purpose or ideological integration or synergy - other than the numerical aggregation of popular votes. The only driving force is the greed to win at all costs with zero attention to integrity and honesty. We should not ever have a Scenario 3.

Here is what we should do:

(1) Stop pre-election coalition alignments completely. Post elections, we can have alignments. The election process will hopefully weed out smaller contestants.

(2) Should coalitions be allowed, an independent body should choose the coalition composition and how many seats each party gets. Make sure parties like Party B in a coalition does not end up with seven seats.

(3) Perhaps a USA style two party system is the answer.

(4) No role in the government (no ministerial posts) for tiny members of coalition alliances. Basically, make sure there is not a huge leverage of numbers.

Long live democracy!!!!


Post a Comment

Links to this post:

Create a Link

<< Home