CPI – will the PBS get it right (ever)?

The Pakistan Bureau of Statistics (PBS) seems to have no response on being pointed out at the faulty calculations of gas price impact on the CPI (see: CPI – correction time for PBS, again! published Aug 5 2019). It has been two days and counting. One wonders if the inaction is because the error, unlike the previous time in November 2018, is on the lower side, as it understates the month-on-month inflation. Whatever it maybe, the silence is disturbing, and surely matters of importance such as inflation numbers, should be given more attention to.

Having erred on the inflation numbers, which stands to understate the year-on-year CPI impact by no less than 71 basis points, one hopes that the rebasing exercise, due to be implemented shortly, covers more than just re-assigning of consumption weights and alteration of the basket. While the existing computation as regards month-month numbers errs on understating, the year-on-year impact is massively overstated, as no heed has been paid on the benefit of previous slab allowed to consumers falling in respective consumption categories, which was not previously the case.

It may be too much to ask for those computing CPI at the PBS to add one more column to the excel sheets, and compute the impact based on weighted averages. But surely, the PBS could still try and assess the actual impact on consumers as a result of gas price revision, which is certainly much less than 142 percent being presently computed, and much more than 31 percent which will be computed come October 2019.

The trick is simple. The CPI is supposed to reflect the changes in price, and the changes in price in case of gas, will only be truly reflected after incorporating the benefit of previous slab. While it may be difficult reaching an exact number, but using ballpark estimate would still be closer to reality than simply ignoring it altogether – even if the weighted average numbers do not come into play.

One also hopes that the PBS comes out of its shell, and corrects the most recent flaw. And it would not take more than just correctly calculating the simple averages. It needs one lesser excel column than weighted average too. Win-win. Also, they would not have to go looking for ‘experts’ to work simple averages out. Surely, the Internet works fine at the PBS offices, Goggle it! .