Per­for­mance Doesn’t Al­ways Pay

Some pa­ram­e­ters of gov­er­nance mat­ter more than oth­ers at the bal­lot box

India Today - - THE STATE OF THE STATES - by Bibek De­broy and Laveesh Bhan­dari

In the re­cent by-elec­tions held in Haryana, Bi­har, Andhra Pradesh and Ma­ha­rash­tra, the Congress got a drub­bing. There is no doubt that per­cep­tions about cor­rup­tion and in­fla­tion in­flu­enced the re­sults. Ques­tion: so how much do de­vel­op­ment and gov­er­nance af­fect elec­toral for­tunes? Data is in­con­clu­sive.

While there is some ev­i­dence of good gov­er­nance be­ing re­warded and bad gov­er­nance be­ing pun­ished, one has to pin down what gov­er­nance means. This con­cerns cit­i­zen ex­pec­ta­tions about a state govern­ment. Ex­pec­ta­tions are about a pack­age that in­cludes an en­abling environment, law and or­der, phys­i­cal in­fra­struc­ture ( bi­jli, sadak and pani), de­liv­ery on so­cial in­fra­struc­ture and re­duc­tion in cor­rup­tion. How­ever, some of these are out­side the pur­vey of state gov­ern­ments, in­tro­duc­ing a vari­able in cal­cu­la­tions.

There have been re­cent statelevel elec­tions in As­sam, West Bengal, Ker­ala, Tamil Nadu and Puducherry. Data is avail­able on party per­for­mances at the level of in­di­vid­ual con­stituen­cies in these five states. With some dis­ag­gre­ga­tion, data on so­cio-eco­nomic in­di­ca­tors can also be cre­ated at the level of con­stituen­cies. We thus know the im­prove­ments (or de­te­ri­o­ra­tion) in so­cio-eco­nomic in­di­ca­tors be­tween 2006 and 2011 and we also know the changes in vote shares of par­ties be­tween 2006 and 2011. Is there any cor­re­la­tion? It is im­por­tant to men­tion that this ques­tion is be­ing asked at the level of the con­stituency, not at the ag­gre­gate level of the state.

The re­sults show that if one dis­ag­gre­gates in this fash­ion, what mat­ters most are in­di­ca­tors that have a strong and di­rect bear­ing on eco­nomic per­for­mance of house­holds— per capita in­come and road con­nec- tiv­ity. So­cio-eco­nomic cri­te­ria such as in­fant mor­tal­ity rate, im­mu­ni­sa­tion and ed­u­ca­tion-re­lated vari­ables do not mat­ter as much. Per­haps not be­cause they are unim­por­tant, but be­cause the elec­torate is aware that these are de­ter­mined more by cen­tral govern­ment schemes. Even crime is not im­por­tant in an over­all sense, but vi­o­lent crimes, eco­nomic crimes and those per­tain­ing to so­cial un­rest have an im­pact.

Does this mean that other so­cioe­co­nomic in­di­ca­tors have no in­flu­ence on elec­toral out­comes? Not quite, be­cause the sam­ple has been re­stricted to As­sam, West Bengal, Ker­ala, Tamil Nadu and Puducherry. In these five states, im­prove­ments in other so­cio-eco­nomic in­di­ca­tors have not had a sig­nif­i­cant sta­tis­ti­cal in­flu­ence on vot­ing pat­terns. There are dan­gers in ex­trap­o­lat­ing this ex­pe­ri­ence to other states.




Newspapers in English

Newspapers from India

© PressReader. All rights reserved.