Elections 2018: Mapping the trends

The basis of any electoral process is an authentic census — the more accurately people can be counted, the more precise is the representation of their aspirations and rights. And while political parties in the wake of the 2018 electoral results have raised a hue and cry about their lost mandate, few realise how little inaccuracies in the census process can have larger, more ground-shaking and shaping outcomes when it comes to the elections.

The sixth population and housing census of 2017 was ultimately carried out in a hurry. Its results were not issued by the last government, because as well-placed sources put it, nobody in the outgoing government wanted to deal with the new social dichotomies that have been thrown up by the census results.

One such phenomenon, for example, is in language data compiled by the Pakistan Bureau of Statistics (PBS) and submitted before the Council of Common Interests (CCI). Census 2017 results reveal a five percent drop in Punjabi speakers in Punjab while also showing the lowest population growth rates. What makes this more interesting is the fact that there is a rise in speakers of other languages — Urdu, Sindhi, Pashto, Balochi, Kashmiri, Seraiki, Hindko and Brahvi.

Then there is the reconfiguration of the urban-rural divide. Rural population in Punjab as a percentage has decreased from 68.7 percent (1998) to 63.14 percent. Simultaneously, urban population in Punjab has increased from 31.3 percent to 36.86 percent. Together, these factors signify how Punjab is slowly but surely turning towards its cities as the engine of growth.

Beyond the glib claims of ‘Naya’ and ‘Purana’ Pakistan, the minutiae of election results throw up some interesting insights into how Pakistan polled.

How does this connect with the elections?

New delimitations or the carving of new constituencies were to ultimately rely on this data to make final electoral maps. Rural areas that now resembled towns, for example, needed to be recognised and rationalised in official documents. That much did not happen in the run-up to the elections as allegations of foul play in the delimitation process continued to be made and, in fact, even taken to the Supreme Court for adjudication.

But perhaps, this is the heart of the matter: electoral numbers issued by the Election Commission of Pakistan (ECP) cannot be completely relied on for an accurate picture since there remain large doubts about the transparency of the electoral process itself. And since the electoral process is premised on the delimitation process, the doubts carry much weight.

The analysis below, however, has been made by setting aside the legitimate question marks over the electoral process. Instead of focusing on who won where and by how much, we have attempted to understand where the close contests eventually came to be and how this compared to the elections of 2013. Similarly the provincial results also offer some intriguing insights. Despite the shadow hanging over a large number of electoral battles there is something to be gleaned from the data, if one looks at some of the trends.

If viewed on a map, the number of close contests — where the margin of victory was less than 10 percent — increased as compared to 2013. Many of these close contests in the last elections were played out in Balochistan, with a few in southern Sindh, eastern Punjab and parts of northern Khyber Pakhtunkhwa. This time, the site of close contests was mainly Punjab.

In total, the number of close contests in the National Assembly this year is 64. This can be further divided by provinces: eight contests in Khyber Pakhtunkhwa (NA-10, NA-11, NA-13, NA-21, NA-22, NA-33, NA-35 and NA-36); two in the tribal areas (NA-42 and NA-49); 39 contests in Punjab (NA-57, NA-67, NA-71, NA-73, NA-74, NA-87, NA-88, NA-89, NA-90, NA-91, NA-94, NA-98, NA-100, NA-105, NA-106, NA-108, NA-110, NA-112, NA-114, NA-117, NA-118, NA-126, NA-129, NA-131, NA-139, NA-140, NA-151, NA-152, NA-157, NA-159, NA-160, NA-161, NA-166, NA-172, NA-174, NA-175, NA-182, NA-187 and NA-188); 12 in Sindh (NA-196, NA-204, NA-212, NA-215, NA-220, NA-221, NA-230, NA-237, NA-239, NA-241, NA-248 and NA-249); and three in Balochistan (NA-259, NA-267 and NA-272).

But what do these close contests indicate?

Contests where the margin of victory was less than 10 percent in 2013 (L) vs 2018 (R) | Dawn GIS
Contests where the margin of victory was less than 10 percent in 2013 (L) vs 2018 (R) | Dawn GIS

Obviously one thing that they indicate is that large parts of the country are politically polarised. But if the data is at least somewhat accurate, it also indicates a rising trend in the popularity of one party or another. Those who weren’t major players the last time round became so this time, either because their personal standing improved or their party’s did.

The lowest margin of winning was in NA-21 (Mardan) with 35 votes; the Awami National Party’s (ANP) Ameer Haider Khan Hoti emerged victorious from this constituency. Another double-digit difference was reported from NA-91 (Sargodha) with Zulfiquar Ali Bhatti of the Pakistan Muslim League-Nawaz (PML-N) winning by a mere 99 votes.

Western Punjab saw a pitched battle between the Pakistan Tehreek-i-Insaf (PTI) and the PML-N. The constituencies of NA-94, NA-98, NA-114, NA-182, and NA-188 all went to the PTI in this clash. The PTI winning (or taking in independents) from western Punjab is also reflected in provincial results from the same region.

Meanwhile, after a very long time, Karachi also saw its share of close contests. Five constituencies are officially earmarked as those where the margin of victory was less than 10 percent — although, arguably, the number of close contests was even higher. While by-elections have already been announced, the true test of these electoral results will be in local government polls. Whether the PTI can replicate its popularity at the local level will ultimately define whether its vote bank in Karachi is genuine or manufactured. — Dawn (EOS).

Leave a Reply

Your email address will not be published. Required fields are marked *