Alert Sign Dear reader, online ads enable us to deliver the journalism you value. Please support us by taking a moment to turn off Adblock on

Alert Sign Dear reader, please upgrade to the latest version of IE to have a better reading experience


Turnout claims are misleading


Your Name:

Recipient Email:

THE large number of people that came out to vote on May 11 is seen as a sign of democracy.

And it seems that this election witnessed a considerably higher voter turnout than past elections. The numbers, undoubtedly, are promising.

Take NA-1 Peshawar where PTI chief Imran Khan trounced ANP’s Ghulam Ahmed Bilour. The Election Commission claims that the turnout in this Peshawar constituency, a city that suffered a number of terrorist attacks in the recent past, was almost twice that of 2008.

According to the ECP website, the turnout in 2008 in NA-1 was 22.97 per cent, while on May 11 it was 46.18.

This means that the turnout increase from 2008 to 2013 is 100 per cent, which is huge.

However, a closer look at the figures shows that the number of registered votes in the constituency in 2008 was actually higher than in 2013. In 2008, 387,083 voters were registered here; by 2013, the Election Commission had revised this down to 320,578, a decrease of 17 per cent in the voters registered for this seat.

The revision occurred due to two reasons — one the Nadra/ECP effort to remove ‘fake votes’ that had entered the rolls prior to the 2008 elections and the addition of new voters who became eligible to vote after the 2008 elections.

However, this exercise means that NA-1 ended up with a lower number of registered voters in 2013 than in 2008. As a result, it is not possible to compare the turnout rate of 2008 with that of 2013.

This is why the claims of the ECP as well as others about a higher voter turnout are misleading at the very least.

Take NA-23, Kohistan. Here the ECP claims that the turnout increased from 17.77 per cent to 40.24. But this too is misleading. In 2008, the number of registered voters here was 317,734, but in 2013, it had come down to 127,028, resulting in a 60 per cent decrease from 2008.

How can the increased turnout of a voter list that is less than half of what it was in 2008 clearly depict the increase in number of votes polled in 2013?

At the same time it is next to impossible to calculate the accurate turnout figures now that the numbers of registered voters have been rectified.

This cannot be done because neither Nadra nor ECP have made public the figures removed and/or added to the electoral rolls. The lack of information on how many votes were fake (and were removed) and how many were new (and were added) between 2008 and 2013 means that the new registered voters numbers on their own can be of no help.

The only way to get some sense of the change in the voter turnout from 2008 to 2013 is to compare the actual number of voters that polled in 2008 and then in 2013 in each of the 272 constituencies.

In this way, it is possible to relate the exact increase in the number of people who turned out to vote.

However, this method assumes that all polled votes in the 2008 elections were genuine, which is also a faulty assumption — if there were fake votes present, chances are some of them would have been polled.

Nonetheless, this method will be a far better indicator of the number of people who voted.

Hence, for the NA-1 case, the increase in the votes polled is 66.5 per cent and NA-23 witnessed a 9.5 per cent decrease in the number of votes polled.

As the table shows, these discrepancies are present in a number of constituencies.

However, this can be rectified if the ECP releases the details of the votes removed from each constituency since the 2008 elections and the ones it added. It should also provide details on how many of the removed votes were duplicate and how many bogus.

This will allow analysts to more accurately determine the polling percentage in 2008 and may make it easier for them to analyse the 2013 results.

DAWN_VIDEO - /1029551/DAWN-RM-1x1

Comments (16) Closed

SHH May 22, 2013 12:57pm
"It is a capital mistake to theorize before you have all the evidence. It biases the judgment." (Sherlock Holmes). This article seems to be suffering from this error. While it rightly poitns out an important aspect of the election data, it is wrong to immediately jump to conclusion and state that ECP claims are misleading. While it is true that it is impossible to have an accurate record of fake voters and bogus votes, and hence to determine the "real" voter turn out, I am sure that the available data can be used to gnerate some very useful statistics on voter participation.
Haseeb May 22, 2013 01:26pm
i don't understand. If this time percentage is higher then it means the turnout is much higher because last turnover was on fake votes.
Interested May 22, 2013 03:15pm
Where is the table?
Usman May 22, 2013 07:18am
Please mention the fake voter figures as well and deduct them from the vote coasted 2008, then result will be much higher than 44%
Mohsin May 22, 2013 07:22am
I think the tone of this article is as (or more ) misleading than the claim of higher voter turnout. The author has failed to recognise and highlight the fact that leaving everything else aside, level of political participation across all strata of society is very positive sign for democracy. To assert that higher voter turnout claim is misleading is not correct. There can be differences in the method and resulting number of how much higher it was, but no one can deny that voter turnout was higher than any previous election. If you consider the corrections in electoral rolls, the nature of corrections includes removing people from multiple addresses and removing people whose record was not found in NADRA. In any event, people who were listed in multiple constituencies could not have casted their votes in all at once. So there were absent from polls anyway. Total number of votes casted is a good indicator .. specially if you normalise it across both elections with percentage of youth becoming eligible for voting and the death rate. Not that difficult to do. It will clearly show you that number of voters casting their votes defiantly increased. It was evident from the queues on the election day. Of course you need to take the number of votes rigged as constant !!
Uzair May 22, 2013 12:22pm
The comparison is very much possible in NA-1 despite the reduced number of registered voters. if we calculate the total number of voters from the figures of total vote and turn out percentages, we find out that in 2008 88.5k votes were polled while in 2013, total number rose to 147k. Hence a total increase by almost 80%. Hence we can consider the results to be very much optimistic. :-)
FJ Siddiqui May 22, 2013 09:04am
For the next elections efforts should be made to reduce the number steps by incorporating some form of electronic identification verification and vote casting. Voting may continue for more than one day as it requires less staff and paper work. As NADRA database is there it is quite a possibility. If election commission start works from now it may be possible in most parts of the country. Where there is a will, there is a way.
Ramem May 22, 2013 11:26am
Its not too difficult to do the math and come up with the correct turnout %rate. Count the total number of votes casted across the country and compare it with the total number of registered voters across the country. You dont need any other data.
A Vetta May 22, 2013 11:45am
My apology, I do not understand this piece. Simply put; turn out = (number of votes cast)/(Total number of votes). Thus, 'turn outs' being ratio statistics are always comparable.
Azmat May 22, 2013 08:07pm
I agree with others here who left comments this article is very misleading. If you comparing turn out rates do not include variables simple rule of math, if you want to introduce than add all variables which you probably don
Allauddin Khan May 23, 2013 02:07am
the position of the writer is quite agreeable because of the loopholes in the two supra authorities from NADRA to ECP. there are many irregularities and almost every educated pakistani konows about this state of dilemma.
Arnold May 23, 2013 06:09am
25 percent of 100 is 25 50 percent of 20 is 10 So, suppose the number of registered voters drops from 100 to 20. If 25 people showed up in the previous elections, the turnout was only 25 percent. If only 10 people showed up in the present elections, the turnout is 50 percent. Do you see that as misleading?
Arnold May 23, 2013 06:11am
I do not understand why there were fake registered voters earlier on if they didn't show up. If they had shown up in the past elections, the turnout in the previous elections would also have been very high in percentage.
Farooq Amin May 23, 2013 06:32am
@ NA-1 - Comeone! Turnout is meaure of % votes casted out of total registered voters. Its not the absolute number of votes and should not be confused with it. ECP states are correct. I didnt read the complete article because writer has no idea about the disctinction between turn out % and absolute vote number. This article is misleading. Lastly 2008 total number of votes were higher because of fake votes as mentioned by writer himself so what to compare?
excalibur May 23, 2013 08:05am
All the detractors should read again. The analysis has merit in it. Why be in self denial ?
Inam ulha May 23, 2013 03:08pm
Y this writer is not talking about HUGE turn out in Karachi. This guy needs some lesson in basic maths.