PG's Blogs


Ghar Wapsi: CEO-KA Style

  • Google search gives about 8,17,000 results for a ghar wapsi (home coming or re-conversion of religion). Member of Parliament Sakshi Maharaj says, ” [it] is no conversion but just a process to guide these people to the faith where they actually belong.” Budhism, Jainism, Sikhism, Islam and Christianity and many other religions all grew by converting believers of other religions. Conversions from these religions would then be ghar wapsi. However, adding fuel to fire, All India Majlis-E-Ittehadul Muslimeen (AIMIM) chief and Hyderabad MP, Asaduddin Owaisi, told in a public speech that every person is born a Muslim and later converted to other faiths. Thus any conversion to Islam would also be home coming.
  • Ghar Wapsi is a hot topic for the past one month even as PK mops up hundreds of crores at box office. OMG!
  • What is not reported is a silent conversion of religious faith of voters as recorded by CEO-KA in electoral rolls, without consent or knowledge of such voters. The process is: A voter needs to edit personal data in the electoral rolls. Innocently he or she uses CEO-KA portal to modify the record. On editing and saving, the record over-writes some other voter’s record. One citizen loses his voting right and the other (editing the data) gets two entries, potentially a bonus vote, violating the democratic principle of one person one vote. CEO facilitates unequal legislative representation to some people. The new record may be of a person of different age, sex, or religious faith (suggested by voter’s name and relative’s name).
  • See my post of 21 May 2013. CEO-KA admitted in April 2013 that about 10,000 voters were deleted and equal number of duplicate records created due to this class of software and process error in Jan and Feb 2013. Though he stated that the records were corrected, in fact they are not. Unfortunately, even after two years of reporting this problem to CEO and ECI in writing and raising in several meetings, the problem persists. Thousands of citizens who lost their voting rights over the years due to this error have no ghar wapsi to the electoral rolls. Rolls published on 05 Jan 2015 smell foul with all too clear indication of the error, possibly corrupting about 200 records out of about 10,000 modifications. See some sample voter records before and after modifications in the table below. Rows with serial numbers prefixed with # show the records before modification, and the next rows show the modified version.
  • When we can detect such errors with software, Electoral Roll Management System (ERMS) could have easily raised alarms when such modifications were made. Special audit trails could have alerted higher authorities about possible errors so that the changes are verified.
  • This is not just an alarmist call. Authorities have earlier accepted that such errors have occurred. I know some friends who have lost their names from the records this way and are not yet restored despite reporting to CEO-KA.

Voter AAA, son of OOO

  • Expectant parents debate over what to name their baby when it arrives. For innovative names, I suggest the electoral rolls of Bangalore. There are about 2,000 records with names like ABC, XYZ, AAA, EEE, OOO, etc. XYZ is the top favorite, with about 400 entries. See the sample images below, snipped from CEO-KA website by searching for voter records.
  • This is not a new error and has been reported to CEO-KA more than an year ago.
  • As per schema published by ECI, relative’s name can be blank in a voter record. There are thousands of records with blank relative names, which is acceptable. But, entering absurd strings in place of a name is not funny.
  • This kind of errors are easy to find and fix. If only CEOs, CEO-KA in particular, care about quality of electoral rolls.

Thank You, and Congratulations, CEO-TN

  • A news story in Hindu of 06 Jan 2015 states “on an experiment, about 58,000 probable duplicate entries in Chennai were spotted. Subsequently, the conventional system of field verification was carried out. Eventually, 15,736 entries were found to be duplicate. “That’s how we removed them from the rolls,” explains Ajay Yadav, Joint CEO (Information Technology)”
  • A CEO-TN press release states, “2,87,598 names were deleted due to shifting, death of electors and duplicate entries. Also, on experimental basis a photographic deduplication software was tried for Chennai Corporation. It resulted in detection of 15736 probable duplicates.”
  • Header line in Hindu states “Novel exercise to spot duplicate voters.” and the byline “Authorities embark on complex initiative.”
  • De-duplication is an activity mandated by ECI before publishing every new version of electoral rolls. It is neither novel nor complex. ECI conducted a workshop of IT officials from CEO’s offices of all states on 24th and 25th May, 2013, at Nainital. They were kind to invite me. In the conference, CEO, Manipur, talked about how she identified duplicate voters by comparing photographs of voters using features of Picasa from Google.
  • I have been reporting suspected duplicate entries in the rolls of various states based on voter record text comparison. Inspired by the talk by CEO, Manipur, I added photo comparison feature to my software. ECI helped me by sending a small sample database dump of Chattisgarh to prove the feature. In Jul to Sep 2013 I demonstrated the software to ECI technical team. The software was installed at ECI and proven. Subsequently, I guess that the software was not used.
  • From 04 Jul 2014 to 26 Aug 2014 officers of CEO-TN (including the Joint CEO) and I exchanged 16 emails. I got a small sample dump of their voter database to run my application to detect duplicates by comparing voter record text, followed by photo comparison. I shared the application with them and helped them run it on their computers. Then there was silence from their end.
  • Whereas CEO-KA has been responding to my observations stating that, “We know the errors in our ERMS and know how to correct them” and do nothing for years, I am glad that CEO-TN has acted to correct the errors in Chennai rolls. New version of the rolls are published on 05 Jan 2015, which I shall analyse to verify the quality.
  • I also hope that other CEOs take lessons from CEO-TN though they have mostly ignored my input for years.

New Electoral Rolls

  • CEOs have published new versions of electoral rolls at their websites today, 05 Jan 2015.
  • In view of assembly elections in Delhi and BBMP elections at Bangalore, rolls of these two cities are important. Please check if your names are in the rolls. Having sighted in the last version or having voted in the last elections do not guarantee that your name would be in the rolls now. Even if you find entry in the rolls, the data could have been corrupted.
  • Despite ECI guidelines about the format of the rolls, various CEOs differ in the format. Delhi CEO has chosen a new format this time – publishing only supplements to the version published on 15 Oct 2014. Thus, to check the roll of a part, we have to open two documents. CEO Karnataka has published single document per part, with main and supplementary part. I wish all the CEOs took one standard approach and stuck to it rather than surprising and confusing the users.
  • Over the next few days I’ll analyse the current version of the rolls and report the main findings here.

Delhi Rolls – Oct 2014

  • Today’s news report states that assembly elections in Delhi are likely to be held by middle of February. The electoral rolls at CEO-DL website were published on 15 October with all the previous errors. CEO-DL and ECI have not acted on the errors despite reporting such errors scores of times. The main issues are
  • 1. More than 11 lakh sets of duplicated voter records. The list can be narrowed down by comparing voter photographs, not available in public documents. I have provided a software application to ECI two years ago to compare photographs, accessing voter database. No actions.
  • 2. ECI recommends maximum 1,400 voters per urban booths. EVMs can register maximum 2,000 votes. Delhi rolls have 2,107 parts with 1,400 voters and 114 with more than 2,000 voters. Analysis of election results of various states show that when the voter counts in a booth increase beyond about 900, voter turnout% drops. It drops sharply beyond 1,600. This is due to physical limitations and not due to voter apathy – the favourite slogan of officials and politicians, insulting the citizens.
  • Repeated mails and calls to politicians also have not stirred any interest in them. My letters to CJI, Minister of Law, and to the Prime Minister have not evoked any response.
  • I’ll analyse the next version of rolls, expected this week, and update here.

ECI’s Election Balls

  • News report states, “The Election Commission on Friday announced single-day poll in Maharashtra and Haryana on October 15, setting the ball rolling …” Election Commission keeps rolling balls without concern for the ground reality about the state of electoral rolls.
  • On 25 April 2014 media published “Election Commissioner HS Brahma has apologised to Mumbaikars for the missing names of voters in the Lok Sabha voter’s list.”
  • While there could be lapses in the system, what is appalling is the contempt for helpful feedback and disregard of ECI rules and guidelines by many CEOs and the ECI.
  • Hand Book for Electoral Registration Officers, Election Commission of India, 2012, Chapter I, paragraph 46, ECI has directed CEO-MH to publish Electoral Rolls of 59 specific constituencies in English. Chapter I, paragraph 41 ibid states, “… at CEO’s website, the roll shall be put in public domain, in a PDF format on the same day when the roll is claimed to be published … The draft roll on the web shall be put in a text mode …” Text in ERO Handbook is clear with words like directive, shall, and should, that it is not left to the discretion of CEO to publish the rolls in English.
  • I have been hoping to analyse the rolls of Maharashtra using sample English versions. Though I wrote to CEO-MH a few times requesting him to publish the rolls of 59 specific constituencies in English, as required by ECI directives, he has chosen to ignore the mails.
  • Based on analysis of only voter counts in various booths of Haryana, the other election going state, it is evident that the CEO has not honored ECI directives.
  • EVMs manufactured since 2006 cannot register more than 2,000 votes. 15 booths in Haryana have more than 2,000 voters. This may be considered illegal because it denies voting rights to people by design.
  • RO Handbook, 2014, suggest maximum 1,200 voters per rural- and 1,400 voters per urban-booth. In Haryana, 4,088 (25.4%) booths have more than 1,200 voters and 1,126 (7%) have more than 1,400 voters. This violates ECI guidelines on booth size.
  • In Haryana we see very sharp decline in voter turnout% when the booth size is larger than 1,600. Unfortunately, CEO-HR has not responded to my mails with this observation. Nor has ECI responded.

A note on NOTA

  • Electors can reject all the candidates by pressing a button marked NOTA (None of the Above) in Electronic Voting Machine. This was introduced for the first time in the recent Loksabha elections after a Supreme Court judgement
  • Like many others, I wondered if people would use this option or abstain from voting when they do not trust any candidate. I also suspected that very few voters would know about and understand this option. Pleasantly, I am wrong!
  • So far only Tamil Nadu has published Form-20, that too only for 27 out of 39 parliamentary constituencies. Form-20 gives details of votes polled by each candidate. Count of NOTA in these constituencies ranges between 4,150 and 46,559; 0.42% to 5% of the total votes polled. The graph below shows the counts of NOTA in different ranges:
  • In 25 out of 27 constituencies, 0.5 to 2% of the total votes polled are NOTA. In Nilgiris, from where Mr. A. Raja of 2G scam contested, 5% of the voters (46,559) have chosen NOTA. This is a significant number. Here, only the winning candidate and the runner-up have got more votes than NOTA. In other constituencies, irrespective of the number of candidates (which varied from 8 to 38 in a constituency), only 3, 4, 5, or 6 candidates have scored more than NOTA. People have clearly rejected the candidates rather than voting undeserving.
  • In the coming elections more people may choose NOTA. Winning margins in Tamil Nadu have been huge in most of the constituencies and NOTA count is not anywhere close to the winning margin. This may not be the case in other states. This may not be the case in future elections.
  • ECI does not have a policy on actions based on such data. When large numbers of citizens reject all the candidates, and if that is not actionable information, meaning of the option dilutes. By exercising the option in the first opportunity, voters have sent a message to the aspirant politicians and the ECI.
  • We hope for some action plan and meaningful action.

ERMS: Sound Guidelines; Insensitive Implementation

  • In earlier blogs I argued that when the number of voters in a polling booth is higher than 1,320, then not all the voters in the booth can cast votes in the allotted polling hours. CEOs deny voting rights of citizens by design when they increase the booth size beyond a limit. This argument was based on logic. With further analysis, we have evidence.
  • After the Assembly Elections 2013, CEO, Karnataka, published Form-20 giving the details of votes polled in each booth of the state. He also published the count of voters in each booth in another document. From these documents I extracted voter counts and votes polled for 41,441 booths. (CEO-KA has later removed these documents from his website. Document properties show the date of creation as 23 May 2013.)
  • We find a clear correlation between decline of voter turnout % and polling booth size whereas the number of people who vote does not increase beyond a count.
  • The table below gives the booth sizes in different ranges, number of booths in that category and turnout in number and turnout % in them. Because a larger % of big booths are in Bangalore, I have given the counts for Bangalore in separate columns.
  • The table above and bar graph below clearly shows decline in voter turnout % when the size of booth crosses 1,400 voters. When the size crosses 1,600 voters, the decline is even sharper. Whereas the turnout % in Bangalore is lower till the booth size is about 1,600 voters, for booths larger than that, the figures are close to those of Karnataka as a whole.
  • Whereas the voter turnout % reduces significantly when the booth size increases beyond 1,600 voters, the counts of people who vote in these booths do not increase much with the size of booth. No correlation of votes polled and booth-size beyond a size indicates that booth’s capacity and not voter apathy reduces voter turnout %. Bar graph below helps in visualising this reality too.
  • Other than some increase at the tail-end, maximum number of voters are found in booths whose sizes are between 1200 and 1400. When the size increases, voters may be de-motivated thinking of possible long queues. This is an assumption based on the data. Surveys with this hypotheses may give better insights.
  • When we talk of the need for electoral reforms and voter apathy, our priorities are wrong. The CEOs should first adhere to the rules and guidelines of ECI and ECI should ensure that they so do. Reforms before that would be like paint on a rusty surface, allowing the rust to eat the metal inside while exhibiting shine outside.

Turn out the voters and cry about poor voter turnout

  • Supplementing the discussion in the last blog, the table below gives the sizes of booths in 6 cities. In a booth with more than 5,500 if more than 20% of the voters cast votes in 11 hours, that is a matter of concern – not poor turnout.
  • Feature of EVM technically puts a limit of five votes per minute. This could be achieved in a simulation, without any gap between two voters and achieve 3,300 votes in the 11 hours duration of voting. Thus even in a simulated run, not all the voters in a booth can vote. Being illogically optimistic, if we consider a gap of 3 seconds between two voters through the day, then we have 4 voters per minute, which allows 2,640 voters to cast votes during the day.
  • However, ECI guideline of maximum 1,200 voters per booth is pragmatic. In the sets of samples, 36% of booths in Delhi, 32% in Chennai, 26% in Bangalore, 16% in Mumbai and 1.4% in Kolkata have more than 1,200 voters. This will certainly reduce the voter turnout, by design and not due to voter apathy as pronounced by armchair experts and authorities. We should be alarmed if we find a high voter turnout in large booths.
  • When ECI cares much for the lone voter in a forest, provisions in several populated areas are grossly inadequate to allow citizens to vote.
  • Bad electoral management can turn voters out and complain about poor voter turnout.

Simple Arithmetic for Election Commission

  • A Times of India story states, “Mahant Bharatdas Darshandas is the lone voter in the midst of Gujarat’s Gir forest, home to the Asiatic lion, for whom an entire election team sets up a polling booth every election — and will do so again on April 30. … a polling team [of 4 to 5 people] travels around 35km to reach the hamlet of Banej inside the Gir forest, located in Junagadh district.” We salute the spirit and efforts of ECI to include every eligible citizen in the democratic process of elections.
  • When we care so much for the lone voter in a forest, are the provisions in populated areas adequate to allow every voter in a booth to cast his vote?
  • “Operationally an Indian EVM is a set of two units – the ballot unit and the control unit. A vote can be recorded only after the presiding officer enables the ballot unit through the control unit. However, even the presiding officer cannot enable the ballot for twelve seconds after every ballot is cast. Thus, a maximum of five votes can be cast in one minute.” – Dr. SY Quraishi, page 192, An Undocumented wonder; The Making of the Great Indian Election, Rainlight/Rupa.
  • Taking minimum 12 seconds per voter, an EVM technically allows casting of maximum 5 votes in a minute in ideal conditions. For each voter to move out of the polling box and the next person to enter, we can consider about 18 seconds. Thus, practically, we expect maximum 2 votes cast in one minute if (a) the voters are enthusiastic and clear about the voting process; (b) the polling staff is efficient and ensure smooth uninterrupted flow of voters; (c) electoral rolls are clean and unambiguous … Remember that the officials search names in paper rolls – not on a computer terminal. People on polling duty take time to find a name in the rolls if the voter does not carry a voter slip with correct details. Though required, CEO-KA has not been giving voter slips in some areas. Queues at times stagnate when a name is not found in the rolls.
  • Polling booths remain open for 11 hours as directed by ECI. With uninterrupted flow of voters and super efficient polling booth staff, 120 votes cast per hour, 1,320 votes may be cast per booth in a day.
  • Paragraph 37 of ERO Handbook, published by ECI in 2012 states, “… Registration of Electors Rules, 1960 indicates that the number of names to be included in any part should not ordinarily exceed 2000. The commission has however, with voter’s convenience in mind, has desired that a part should not have more than 1200 electors in urban area and 1000 electors in rural area.”
  • ECI guidelines are not only for voter convenience, but are necessary to allow the system to work even in ideal conditions. If the voters in a booth exceed 1,200, many voters may not get their turn to vote.
  • Reality: 2,028 booths (more than 26%) out of 7,712 booths of Bangalore have more than 1,200 voters each. Booth# 280 of Sarvagnanagar constituency has 3101 voters. How can we expect more than 50% voter turnout in this booth?

Some Update

  • 1.Moneylife has published an article I wrote summarising the core problems with electoral rolls of Bangalore.
  • 2. Deccan Herald reports, “He [Dr. Quraishi] said the electoral roll in Bangalore was always in a mess despite the City being the IT Capital of India. The CEC had to come down to the City to put the roll in order and to give a dressing-down to several people for the mess. He reminded the voters of Bangalore that the poor voter turnout always favours criminals, given the fact that 30 per cent of the politicians fielded this time are criminals.”

Lean Electoral Rolls of Maharashtra

  • In a statement to Moneylife Chief Electoral Officer of Maharashtra has defended mass deletion of 50 lakh voters from electoral rolls of Maharashtra,telling that it was publicised and the voters should have objected if there were incorrect deletions.
  • ERO Handbook, published by ECI, states “… before taking any action for amendment or transposition or deletion of an entry on the ground that the person concerned has ceased to be ordinarily resident in the Constituency or that he is otherwise not entitled to be registered in the Electoral Roll of that Constituency, the ERO shall give the person concerned a reasonable opportunity of being heard in respect of the action proposed to be taken in relation to him… It is very important that proper enquiry and verification should be made before deleting a person’s name from the electoral roll. The Officer ordering deletion should personally satisfy himself that the deletion is justified because the right to vote, which is a statutory right, is taken away by such deletion. ”
  • I suspect if CEO-MH has taken such due care before deleting the voters. If so, it would be good to see evidences of such actions. In the process of justifying his actions, CEO-MH seem to have forgotten the spirit of “Every eligible person shall be on the roll and name of every ineligible person shall be removed.” stated by CEC on 11 June 2012, as published at ECI website. Unfortunately we seem to have lakhs of eligible voters wantonly excluded from the rolls and lakhs of fake entries and duplicate records entering the rolls.
  • Deleting a voter without due diligence and then expecting the voters to object is like harassing an innocent by prosecuting and then making him to prove that he is not guilty.

Joke of the Year

  • Chief Electoral Officer of Karnataka is quoted in Bangalore Mirror, stating, “…the responsibility in making an error free and clean electoral roll lies more with voters.“
  • I hope this was only a joke from CEO-KA.

BBMP: Waste Management and Electoral Roll Management

  • Dakshina Kannada had 76.6% turnout as per news reports.
  • The same CEO, but a different DEO and EROs. CEO-KA is unable to get good work from BBMP staff. The rolls maintained by BBMP have large % of voters who have moved out of station or out of the world. It has about 20% suspected duplicate entries. We can further prove it if we get voter records with photographs. Also, because of wrong entries, people are turned away from booths – well, we have an example when a lady was turned away because her sex is shown as male, husband, who is 10 years older is shown as her father, though it may be a stray incident. You would understand the reason why I have been reporting impossible age differences between parent and child in electoral rolls. Such smells help the organisation to validate records. Registration process is unfriendly.
  • If we can get English version of Dakshina Kannada rolls, we can analyse the difference in quality and understand why the turnout is better.
  • BBMP’s electoral roll management is not better than their waste management. Do they find some difference between solid waste and an elector?

Voters Turned-out

  • I gave the following text to Times of India, which is also in a mail I sent to ECI and CEO-KA:
  • Typical reactions about reportedly poor voter turnout in Bangalore on 17 April 2014 have been:
  • 1.People who have migrated to Bangalore do not participate in the city’s activities.
  • 2.Being sandwiched between holidays, many people have gone out of station.
  • Most of the people working for social causes in Bangalore have come from outside Karnataka. Turnout is not lower because of cosmopolitan nature of Bangalore. My sample checks show that not many people have missed voting because of holidays.
  • Karnataka has the worst Electoral Roll Management System (ERMS) among the 13 states/UTs I have analysed. I do not trust CEO-KA numbers because the basis on which they are calculated are wrong. Take an example quoted by Anand Yadwad. He checked 20 houses in his complex and found the following flaws in the electoral rolls:
  • 15% of entries were duplicated
  • 33% of people had moved out
  • 5% had expired
  • Wrong data in the rolls also lead to disenfranchisement. E.g., voter list AC1600169.pdf at CEO website shows serial 114, L.C.N. Velu, EPIC# BCW6606073, as male. She is a lady. The polling booth officials did not allow her to vote. We have been reporting thousands of such absurd errors for years, which CEO-KA ignores.
  • The sample check above shows 48% genuine voters in the list. Not all the parts are as bad. However, our electoral rolls have lakhs of fake entries while genuine voters are deleted regularly. CEO-KA and ECI are deaf to regular feedback with evidences. CEO-KA is ridden with very poor quality ERMS software and non-adherence to processes specified by ECI.
  • Using simple technology, we can improve the quality of Electoral Rolls and also significantly reduce human efforts. CEO-KA has so far ignored feedback and rejected offers of help.
  • Turnout calculated based on these inflated rolls are expectedly low, demoralising citizens and leading to wrong actions. In reality, voter turnout was much higher. Apathy is not with the voters, but with the authorities.
  • TOI printed an opposite opinion on people, stating wrongly, “people who have migrated to Bangalore do not participate in large numbers in the cities activities and others are holidaying”. Based on my interactions with various social groups, I think that people who have migrated to Bangalore are more active in democratic duties and social issues than those who are originally from Karnataka. You can see the contradictions in the first and second paragraphs published in TOI.

Election Day: Who’s Apathy?

  • Seeing long queues at polling booths at 0700 is gratifying. By 0715 I got a few calls from young people that they have already cast their votes.
  • I visited a few booths within 5 km from my home. The political party representatives stationed near polling booths help people to find their names in the rolls and give voter slips. Unfortunately most of them have old version of voter lists, published on 06 Jan 2014. About 6 lakh voters have been added to electoral rolls of Bangalore since then, whose names appear in the lists published on 26 Mar 2014. When the party representatives do not find the names in their old lists, they tell the citizens that the names are not in the lists though in fact they could be in the current lists. The reps are ignorant.
  • When the voters go to the booths, the officials there ask for voter slip given by political parties, which makes it easy to verify entries in electoral rolls.
  • ECI letter 491/SVEEP/09/2013-KT(EGS) dated 28 Feb 2013 required CEO, Karnataka, to issue voter slips to individual voters during the Assembly Elections in May 2013. This was done only partially. Recently, CEO-KA announced that voter slips with photographs would be given to individual voters for the Loksabha Elections. I do not know of anyone who has got such slips.
  • In the absence of official voter slips from CEO organisation, booth officers asking for voter slips, and political parties referring to old lists and misleading the voters – it will still be voter apathy, officially speaking.


If you have any questions you may write a mail to us directly and we will answer.

Test Data