CPI rationalized — now methodology’s turn

The Pakistan Bureau of Statistics (PBS), in a rare move, adjusted the monthly consumer price inflation for October 2018. The downward revision is based on the criticism of computing the gas price increase at 104 percent. The PBS, in its defence, was limited by the 5-slab methodology, and it is difficult adjusting the models when the number of slabs changes from five to seven. It is heartening to see an institution of PBS’s stature, open to criticism and making the required due correction.

The PBS is given the benefit of doubt on its limitations regarding changing the methodology midway, when the new rebasing is just around the corner. But one must keep in mind that the “rationalized” CPI has not entirely done justice to the cause. So, while no one is expecting the Bureau to alter its computation methodology just now, it is now highly expected of them to change it, once the base is shifted. For one simple reason – that the methodology has one glaring errors too many to let go of. (Read: PBS owes a correction, published on Nov 5, 2018 & High inflation, low common sense, published on Nov 6, 2018).

The saga of natural gas price and CPI is presented as a test case here. The CPI measures urban inflation with respect to five different income groups (quintiles). Each quintile is assigned a weight, which shows the percentage of population in a specific income group. The highest income quintile (Q5) has the highest weight of 33 percent as per Household Integrated Economic Survey (HIES) 2007-08, which is also the base year of CPI computation basket. It therefore, makes sense, that a higher price increase for that very segment will have a higher impact on overall inflation.

The household expenditure is also based on a detailed survey i.e. the Family Budget Survey. The HIES also records household consumption expenditure across various items, of which lighting is one (see table). So while the PBS has all the relevant datasets with respect to consumption of each income group across various consumption items – it is unfathomable to use simple averages.

The recent gas price revision, even if based on the latest HIES, that of 2015-16, for argument’s sake – should not result in a revision of more than 28.47 percent. The error here has been deliberately kept on the higher side, by negating cylinder gas prices, which if included, would result in a much lesser price increase. The point is that the PBS does not lack data. It just needs to apply it more effectively.

The PBS, unlike many state institutions, is receptive of criticism and suggestions. The criticism that the current methodology is flawed has been well received. The suggestion that it should change for the better, will hopefully get reception too. Especially, when there is no dearth of good quality detailed data, that too, gathered all by the PBS itself.