High inflation; low common sense

Pakistan is not exactly known for sophisticated data availability. The recently released monthly inflation numbers for October 2018 have created quite an uproar in circles keeping a close eye on numbers. The issue in question is the revision in gas price index by a mammoth 105 percent. The Pakistan Bureau of Statistics (PBS) has simply based the mechanism on simple average - negating consumption patterns and weightages. (Read: Inflation: The PBS owes a correction, published Nov 5, 2018).

At first glance, it looked an error. Some thought that the PBS may come up with a correction. But there is actually method to the madness. Tracking previous instances of notified price increase in gas and electricity tariffs reveals that the PBS simply relies on simple averages across various slabs.

Here is a hypothetical scenario. Had Ogra added one more slab for consumption of say over 1000 cubic meter - and raised the price by 300 percent. The PBS methodology would have swiftly disregarded the fact that slab does not even constitute 0.5 percent of total consumers - and the gas index revision would have been north of 200 percent.

Thankfully, previous instances of gas price revision were by and large reflected fairly - not because the PBS back then had a more believable computation methodology - but just because the price increase in those instances was rather even, without new slabs.

Some may worry how the PBS would compute the electricity tariff in the November review of monthly prices. Thankfully, the notified increase does not exceed 15 percent for top most slab - and remains unchanged for nearly 70 percent consumers. The simple average formula would yield a revision of 4-5 percent in power prices. This again, would be only because the price revisions are not bordering on the extreme.

Had there been a new slab with a massive increase - the November inflation would have also been skewed on the higher side - for all the wrong reasons. The gas bit has already been done, and it will be a permanent fixture in the ensuing months - as the base has been set higher.

One wonders why can't the Bureau with extensive data on its disposal, from household consumption patterns broken down to details of income quintile - simply use consumption weightages. Especially, when it is about dealing with publicly available notified prices. Surely, the treatment of other price items, would have more questions than answers.

And go no further than the house rent index itself, on a classic example of an unfair methodology. Recall that the house rent has the single highest weight of 21.8 percent in the CPI basket. Imagine there were only two cities in Pakistan; Karachi and Dera Ismail Khan, with current population. The CPI house rent index would have read 9.5 percent increase - despite Karachi's increase being 0.64 percent. The house rent index revision is also based on simple average of 40 cities.

Surely, using common sense or census data or both would provide a truer picture of affairs. As things stand today, a massive proportion of CPI price computation methodology, is not only unfair, but borders on the ridiculous. It is time, the PBS brings to use its own data sets from census and household consumption surveys. Or else, CPI would keep portraying the wrong picture. Incorrect diagnosis leads to wrong prescription. Pakistan's economy is ailing already. At least get the diagnosis right. Now.