How reliable is Bangladesh’s inflation data?
Economists use inflation data to predict the course of the economy and recommend policies that might move the economy towards the desired path. That begs the question, how reliable is Bangladesh’s inflation data and what’s behind the delays in releasing the time-sensitive data
There seems to be a hush-hush atmosphere regarding Bangladesh's inflation figures, especially surrounding those for the months of August and September.
On 5 October, Finance Minister AHM Mustafa Kamal announced that the rate of inflation (CPI) for August and September were 9.5% and 9.1% respectively, which was the highest in 11 years.
Bad enough as it is, a lot of people - ranging from ordinary citizens to veteran economists - seem to doubt the authenticity of these inflation figures, especially given the Bangladesh Bureau of Statistics' (BBS') reluctance and uncharacteristic delay in releasing the CPI and inflation data for August, as well as, the 42.5 and 51.6% rise in diesel and octane prices, respectively, in August.
But not many are willing to voice their opinions in public.
It is true that octane and diesel are only two of the 422 products included in the consumer basket used to calculate the Consumer Price Index (CPI), and even a substantial increase in their prices may not push inflation up too high. That being said, increases in fuel prices likely had a multiplier effect on the prices of other goods, as most producers and suppliers had to incur higher transportation costs.
To be fair, it is entirely possible that the BBS calculation of the CPI is correct and authentic. But the bureau's lack of transparency in sharing the inflation figures, CPI report and raw price data raise concerns. The simple question remains: if there is nothing to hide, why is the BBS so reluctant in sharing the data?
The BBS usually declares the inflation figures within one week of the following month. That is, the inflation figure for August should have been released by the first week of September and that of September, within the first week of October. But it was delayed by more than a month.
The BBS reportedly prepared the inflation data for August and September in due time and submitted it to the Planning Ministry. But the data would not be released until about two months later, after receiving the green light from the government.
More importantly, any declaration of official inflation figures should be associated with detailed CPI data for the month, which can be used to cross-check the figures for inflation, food inflation and non-food inflation. The BBS, however, does not release the raw data for calculating CPI, blocking any verification by third-party reviewers.
When asked about why the detailed price data for all 422 products were not made publicly available, Najmul Hossain, Deputy Director at Bangladesh Bureau of Statistics (BBS), said "There are security concerns with sharing the raw price data."
Another joint director refused to comment.
However, security concerns make no sense. Calculating or cross-checking the CPI figures only require price data from various regions of the country, which require a lot of resources to accumulate and no third-party reviewer would have the capability to amass this amount of data on a regular basis.
This process does not require data on individual mobile phone numbers, NID or passport information, or anything like that. Even the United States, the country most obsessed with so-called security concerns, makes its CPI data publicly available on the Bureau of Labour Statistics website.
That being said, if the comment is recontextualised in terms of the political sensitivities, their so-called security risks can start to make sense.
Even the CPI report for August 2022 - comprising of its CPI figures disaggregated by region, division and different groups of consumer goods - was released about three months later, on 24 October late at night after office hours, when the entire country was grappling with Sitrang and the subsequent chronic load-shedding, as well as internet blackout.
Many prominent economists had already raised concerns that the BBS might be readjusting or editing the CPI data to fit the inflation rate declared by the Finance Minister, and it might be released when it is no longer possible to cross-check the data or the inflation figures in the market.
Ahsan Mansur, Executive Director of the Policy Research Institute said, "The reluctance to release CPI data raises concerns regarding transparency on the part of BBS and they might be involved in editing and manipulating the actual figures."
Even giving the benefit of the doubt to the government's premier statistical bureau, this practice of releasing the data two months later could at best be considered gross negligence or incompetence of the people involved. Moreover, releasing the inflation data after two months makes the information redundant, as it cannot be used to post-facto, readjust the course of the economy.
As Dr Selim Raihan, Executive Director of the South Asian Network on Economic Modeling (SANEM), said, "While I discourage speculations regarding the authenticity of the data, inflation and CPI data should not have been delayed this much. Economists use inflation data to predict the course of the economy and recommend policies that might move the economy towards the desired path. If the inflation data is delayed, it is difficult to achieve that goal."
But there are more ways in which the CPI report as well as inflation figures released by the BBS has to evolve.
Firstly, the inflation and CPI figures are calculated from the weighted average of the prices of all 422 products included in the consumer basket. As economists have repeatedly complained, the basket remains outdated and is unilaterally applied to calculate inflation for all income groups.
"In a recent study from Sanem, we showed that the inflation figures for low-income households were drastically higher than the official inflation figures," said Dr Selim Raihan.
Different income groups prioritise different commodities and while the BBS does disaggregate inflation data across regions and divisions, it does not do so for different income groups. Hence, inflation figures hardly allow us to capture the real condition of the economy which may have affected the poor and the marginalised disproportionately.