The Consumer Vocabulary / Explanation for Anti Incumbency

The Economic Times - May 2004

Ever since election results were out, there has been an avalanche of comment by politicians, political watchers, journalists, pollsters, intellectuals, advertising and marketing experts and more, analysing the who - what - why of the results. But conspicuous in its absence has been the voice of the voter. "What", some might say in horror. "This whole thing is about the voice of the voter". But as I said in as earlier article "Let's have real vox pop please", in a business we would insist that market share and sales data numbers, especially when unexpected, be explained through the lens of individual customer attitude and behaviour i.e. what customers did, and what they thought, which made them do what they did. The election result is like market share and sales data. Real explanations are what we are yet to get - too few facts straight from the horses mouth i.e. from the people that actually voted, on why they did what they did, and too much conjecture from the rest of us. Confusing, because every explanation provided by some one of repute has been contradicted by someone else of equal repute!

I have three points to make (fools rushes in where angels fear to tread!) (1) there is a lot of informed conjecture and not enough consumer generated fact about what goes on in side consumers (voters) heads. Because there are so may diverse explanations, all of which make so much sense and come so well recommended, we need to actually go back and invest time and money in understanding this verdict better (2) Understanding it better requires two perspectives - a 'supply side / product wise / detailed sales analysis of who bought and dumped whom in which part of the market. And a fact based, formal, explanatory, customer insight on "why they did what they did", obtained by viewing it all through the eyes of the customer (voter). Lots of supply side / product window/ sales and buyer profile analysis has been done. Very little customer insight is available, even though politicians and pundits say they know it all (and contradict each other totally) (3) Wherever we have tried to understand voter thinking better, the measuring instruments have been inappropriate. (quicky type, slap dash, scales used to optimize speed not insight, not in consumer vocabulary, asking for judgement on events rather than on feelings about things, supplier agenda driven questions, not consumer hypotheses driven questions).

Too much conjecture: Consider just a sprinkling of recent opinion, from impeccable sources: "Many of the other explanations for an unexpected electoral result do not stand scrutiny. It is argued in somewhat facile fashion, for instance, that this is a vote where Bharat is sending a message to those enjoying that part of India which is Shining; but then the NDA has been swept out in both Mumbai and Delhi, while the BJP has done very well in many backward rural constituencies.... there is only one way to interpret what the voters have said: they want change and will not be easily satisfied" / "While the rich got richer.... the fortunes of the poor declined year by year. The gulf between India's rich and poor has never looked wider than it does today and the government has fallen into the chasm" / "voter arrogance" / "a somewhat inchoate expression of frustration with the existing order" /a "dramatic reassertion of the will of the mass of Indian people faced by a combination of feudal, dictatorial and neo fascist structures of governance". I think we need to get the opinion of the guys that did it, don't you?

Lots of supply side analysis:

  • Detailed analysis has been done on increase or decrease in vote shares and corresponding seat shares and repeatedly thrown up the small gap in percentage of votes for the victor and vanquished; related analysis has been done on the alliance effect.

  • Some hypotheses have been tested with macro data quite competently. For example, the thesis that reforms have not had a human face has been contradicted by data that they do. However whether this macro data presented as improved frequency distributions conceals the story of anger from a certain section of voters is not known. Calibrating macro level economic data to micro level human feelings data needs to be done.

  • Then we have had lots of "who (gender, income, age, community etc) voted which way" analysis. Some of it has been based on assumptions like Urban vs Rural = rich vs. poor, but that line of thought has been abandoned. The only source of that data on this is an exit poll; and if the exit polls went wrong on account of non representative sampling or the 'lying factor' of people too scared to tell the truth (and not on account of votes to seats projections), then the data is not useable. In fact no post poll data on this count is useful, unless it is tested for correctness on actual vs. survey observed percentage of votes polled.

The consumer insight behind anti incumbency:

Anti incumbency seems to be a catch all phrase that needs more explanation. Does it mean JLT (why not, just like that) or is it "customer dissatisfaction"? How come it applies to Chandrababu but not to Laloo and Naveen Patnaik? And West Bengal? Obviously there is some form of customer satisfaction here. Are the dimensions of satisfaction and dissatisfaction the same on all counts? Yes, there are several specific explanations that have been advanced for each state. But are they customer generated, formally measured? I think we still need to do that, and with a consumer empathetic set of questions. A lot of poll diagnostic questions insist that voters think and talk using pollsters' mental models instead of the other way round. Maybe a good dose of serious qualitative research before framing questionnaires would help a lot.

The National Election Study 2004, a post poll study done by the CSDS team reveals high satisfaction with the NDA government on curbing corruption (31% say improved, 78% say improved or same), security of the country (43%, 78%),India's image in the world(43%,68%), Hindu Muslim brotherhood (34%, 62%), and development of the country (45%,78%). But on umemployment the numbers are 17% saying improved, 44% improved or same, and a whopping 47% saying got worse (compared to 12 to 22% on other parameters). So was that the mandate? Give us more jobs? Was that the face of anti incumbency? And is it just the poor? Also is this about half the GDP of rural India having moved away from agriculture, without a corresponding shift in employment?